hr analytics research

Human Resource (HR) analytics

Learn what is HR analytics and how does it work. Discover examples, metrics and learn how it can be implemented and used in your organization.

Ivan Andreev

Demand Generation & Capture Strategist, Valamis

December 20, 2021 · updated July 31, 2024

13 minute read

What is HR analytics?

Why is hr analytics needed, examples in hr analytics, how does hr analytics work, examples of hr analytics metrics, pros and cons of hr analytics, predictive hr analytics.

HR analytics is the process of collecting and analyzing Human Resource ( HR ) data in order to improve an organization’s workforce performance. The process can also be referred to as talent analytics, people analytics, or even workforce analytics.

This method of data analysis takes data that is routinely collected by HR and correlates it to HR and organizational objectives. Doing so provides measured evidence of how HR initiatives are contributing to the organization’s goals and strategies.

For example, if a software engineering firm has high employee turnover, the company is not operating at a fully productive level.

It takes time and investment to bring employees up to a fully productive level.

HR analytics provides data-backed insight on what is working well and what is not so that organizations can make improvements and plan more effectively for the future.

As in the example above, knowing the cause of the firm’s high turnover can provide valuable insight into how it might be reduced. By reducing the turnover, the company can increase its revenue and productivity.

Read: How to Successfully Implement Learning Analytics in Your company

Most organizations already have data that is routinely collected, so why the need for a specialized form of analytics? Can HR not simply look at the data they already have?

Unfortunately, raw data on its own cannot actually provide any useful insight. It would be like looking at a large spreadsheet full of numbers and words.

Without organization or direction, the data appears meaningless.

Once organized, compared and analyzed, this raw data provides useful insight.

They can help answer questions like:

• What patterns can be revealed in employee turnover?
• How long does it take to hire employees?
• What amount of investment is needed to get employees up to a fully productive speed?
• Which of our employees are most likely to leave within the year?
• Are learning and development initiatives having an impact on employee performance ?

Having data-backed evidence means that organizations can focus on making the necessary improvements and plan for future initiatives.

With the ability to answer important organizational questions without any guesswork, it is not surprising that many businesses using HR analytics are attributing performance improvement to HR initiatives.

Use learning data to accelerate change

Understand learning data and receive a practical tool to help apply this knowledge in your company.

How can HR Analytics be used by organizations?

Let’s take a look at a few examples using common organizational issues:

1. Turnover

When employees quit, there is often no real understanding of why.

There may be collected reports or data on individual situations, but no way of knowing whether there is an overarching reason or trend for the turnover.

With turnover being costly in terms of lost time and profit, organizations need this insight to prevent turnover from becoming an on-going problem.

HR Analytics can:

• Collect and analyze past data on turnover to identify trends and patterns indicating why employees quit.
• Collect data on employee behavior, such as productivity and engagement, to better understand the status of current employees.
• Correlate both types of data to understand the factors that lead to turnover.
• Help create a predictive model to better track and flag employees who may fall into the identified pattern associated with employees that have quit.
• Develop strategies and make decisions that will improve the work environment and engagement levels.
• Identify patterns of employee engagement , employee satisfaction and performance.

Exit interview template

Conduct an exit interview and collect valuable information that can help improve the work culture in your organization.

2. Recruitment

Organizations are seeking candidates that not only have the right skills, but also the right attributes that match with the organization’s work culture and performance needs.

Sifting through hundreds or thousands of resumes and basing a recruitment decision on basic information is limiting, more so when potential candidates can be overlooked. For example, one company may discover that creativity is a better indicator of success than related work experience.

• Enable fast, automated collection of candidate data from multiple sources.
• Gain deep insight into candidates by considering extensive variables, like developmental opportunities and cultural fit.
• Identify candidates with attributes that are comparable to the top-performing employees in the organization.
• Avoid habitual bias and ensure equal opportunity for all candidates; with a data-driven approach to recruiting, the viewpoint and opinion of one person can no longer impact the consideration of applicants.
• Provide metrics on how long it takes to hire for specific roles within the organization, enabling departments to be more prepared and informed when the need to hire arises.
• Provide historical data pertaining to periods of over-hiring and under-hiring, enabling organizations to develop better long-term hiring plans.

Onboarding checklist

Ensure a smooth and effective onboarding process for your new hires and set them up for success.

Understanding the process of HR analytics

HR Analytics is made up of several components that feed into each other.

• To gain the problem-solving insights that HR Analytics promises, data must first be collected .
• The data then needs to be monitored and measured against other data, such as historical information, norms or averages.
• This helps identify trends or patterns. It is at this point that the results can be analyzed at the analytical stage.
• The final step is to apply insight to organizational decisions.

Let’s take a closer look at how the process works:

1. Collecting data

Big data refers to the large quantity of information that is collected and aggregated by HR for the purpose of analyzing and evaluating key HR practices, including recruitment, talent management , training, and performance.

Collecting and tracking high-quality data is the first vital component of HR analytics.

The data needs to be easily obtainable and capable of being integrated into a reporting system. The data can come from HR systems already in place, learning & development systems, or from new data-collecting methods like cloud-based systems, mobile devices and even wearable technology.

The system that collects the data also needs to be able to aggregate it, meaning that it should offer the ability to sort and organize the data for future analysis.

What kind of data is collected?

• employee profiles
• performance
• data on high-performers
• data on low-performers
• salary and promotion history
• demographic data
• on-boarding
• absenteeism

2. Measurement

At the measurement stage, the data begins a process of continuous measurement and comparison, also known as HR metrics.

HR analytics compares collected data against historical norms and organizational standards. The process cannot rely on a single snapshot of data, but instead requires a continuous feed of data over time.

The data also needs a comparison baseline. For example, how does an organization know what is an acceptable absentee range if it is not first defined?

In HR analytics, key metrics that are monitored are:

Organizational performance Data is collected and compared to better understand turnover, absenteeism, and recruitment outcomes.

Operations Data is monitored to determine the effectiveness and efficiency of HR day-to-day procedures and initiatives.

Process optimization This area combines data from both organizational performance and operations metrics in order to identify where improvements in process can be made.

Here are some examples of specific metrics that can be measured by HR:

• Time to hire – The number of days that it takes to post jobs and finalize the hiring of candidates. This metric is monitored over time and is compared to the desired organizational rate.
• Recruitment cost to hire – The total cost involved with recruiting and hiring candidates. This metric is monitored over time to track the typical costs involved with recruiting specific types of candidates.
• Turnover – The rate at which employees quit their jobs after a given year of employment within the organization. This metric is monitored over time and is compared to the organization’s acceptable rate or goal.
• Absenteeism – The number of days and frequency that employees are away from their jobs. This metric is monitored over time and is compared to the organization’s acceptable rate or goal.
• Engagement rating – The measurement of employee productivity and employee satisfaction to gauge the level of engagement employees have in their job. This can be measured through surveys, performance assessments or productivity measures.

3. Analysis

The analytical stage reviews the results from metric reporting to identify trends and patterns that may have an organizational impact.

There are different analytical methods used, depending on the outcome desired. These include: descriptive analytics , prescriptive analytics , and predictive analytics .

Descriptive Analytics is focused solely on understanding historical data and what can be improved.

Predictive Analytics uses statistical models to analyze historical data in order to forecast future risks or opportunities.

Prescriptive Analytics takes Predictive Analytics a step further and predicts consequences for forecasted outcomes.

Examples of analytics:

Here are some examples of metrics at the analytics stage:

• Time to hire – The amount of time between a job posting and the actual hire is a metric that enables HR to gain insight into the efficiency of the hiring process; it prompts investigation into what is working and what is not working. Does it take too long to find the right candidate? What factors could be impacting the result?
• Turnover – Turnover metrics that indicate the rate at which employees leave the organization after hire can be analyzed to determine what specific departments within the organization are struggling with retention and the possible factors involved, such as work environment dissatisfaction or lack of training support.
• Absenteeism – The metric indicating how often and how long employees are away from their jobs as compared to the organization’s established norm could be an indicator of employee engagement. As absenteeism can be costly to the productivity of an organization, the metric enables HR to investigate the possible reasons for high absence rates.

4. Application

Once metrics are analyzed, the findings are used as actionable insight for organizational decision-making.

Examples of how to apply HR analytics insights:

Here are some examples of how to apply the analysis gained from HR analytics to decision-making:

• Time to hire – If findings determine that the time to hire is taking too long and the job application itself is discovered to be the barrier, organizations can make an informed decision about how to improve the effectiveness and accessibility of the job application procedure.
• Turnover – Understanding why employees leave the organization means that decisions can be made to prevent or reduce turnover from happening in the first place. If lack of training support was identified as a contributing factor, then initiatives to improve on-going training can be put together.
• Absenteeism – Understanding the reasons for employee long-term absence enables organizations to develop strategies to improve the factors in the work environment impacting employee engagement.

HR analytics is fast becoming a desired addition to HR practices.

Data that is routinely collected across the organization offers no value without aggregation and analysis, making HR analytics a valuable tool for measured insight that previously did not exist.

But while HR analytics offers to move HR practice from the operational level to the strategic level, it is not without its challenges.

Here are the pros and cons of implementing HR analytics:

• More accurate decision-making can be had thanks to a data-driven approach, which reduces the need for organizations to rely on intuition or guess-work in decision-making.
• Strategies to improve retention can be developed thanks to a deeper understanding of the reasons employees leave or stay with an organization.
• Employee engagement can be improved by analyzing data about employee behavior, such as how they work with co-workers and customers, and determining how processes and environment can be fine-tuned.
• Recruitment and hiring can be better tailored to the organization’s actual skillset needs by analyzing and comparing the data of current employees and potential candidates.
• Trends and patterns in HR data can lend itself to forecasting via predictive analytics, enabling organizations to be proactive in maintaining a productive workforce.
• Many HR departments lack the statistical and analytical skillset to work with large datasets.
• Different management and reporting systems within the organization can make it difficult to aggregate and compare data.
• Access to quality data can be an issue for some organizations who do not have up-to-date systems.
• Organizations need access to good quality analytical and reporting software that can utilize the data collected.
• Monitoring and collecting a greater amount of data with new technologies (eg. cloud-based systems, wearable devices), as well as basing predictions on data, can create ethical issues.

Predictive Analytics analyzes historical data in order to forecast the future. The differentiator is the way data is used.

In standard HR analytics, data is collected and analyzed to report on what is working and what needs improvement. In predictive analytics, data is also collected but is used to make future predictions about employees or HR initiatives.

This can include anything from predicting which candidates would be more successful in the organization, to who is at risk of quitting within a year.

How does it work?

Advanced statistical techniques are used to create algorithmic models capable of identifying trends and future behaviors. These future trends can describe possible risks or opportunities that organizations can leverage in long-term decision-making.

Predictive HR examples

Let’s take a look at how predictive analytics can be used:

Turnover With predictive analytics, an algorithm can be devised to predict the likelihood of employees quitting within a given timeframe. Being able to flag which employees are at risk enables organizations to step in with preventative measures and avoid the cost of losing productivity and the cost of re-hiring.

Organizational Performance Historical data can pinpoint reasons for poor performance, but predictive analytics can make predictions about what initiatives are most likely to improve performance. If engagement levels are identified as being correlated with performance, then organizations can implement specific initiatives that boost employee engagement.

The benefits and challenges of predictive HR analytics

Benefits: Predictive HR analytics enables organizations to become proactive in their use of data.

Instead of fixing past problems, organizations can create a future that prevents problems and solves future challenges before they even happen. This can save on future costs, both in revenue, goals, and productivity.

Challenges: Predictive HR analytics requires a level of skill, technology and investment that many organizations do not yet have.

Many factors also need to be taken into consideration in order to make predictions about employees or potential candidates.

Human beings can be unpredictable and have different personalities, backgrounds and experiences. Slotting people into a black and white algorithm in order to make predictions about their job performance or future poses not just a risk, but an ethical question.

You might be interested in

Key areas of employee development

Learning and development fundamentals

The meaning of Learning Management System (LMS)

Access to 13 certificate programs,
courses and all future releases

Personal Coaching and Career Guidance

Community and live events

Resource and template library

• 4 Types of HR Analytics...

4 Types of HR Analytics [With Examples]

What is HR analytics?

• Descriptive analytics: Analyzes historical data to understand what has happened in the workforce over a specific period.
• Diagnostic analytics: Examines data to understand the causes of past events and behaviors within the HR domain.
• Predictive analytics: Uses statistical models and forecasts to predict future HR events and employee behaviors based on current and historical data.
• Prescriptive analytics: Provides recommendations on how to handle future situations and challenges in HR by analyzing potential outcomes and scenarios.

Benefits of HR analytics

• Practicing evidence-based HR : Making decisions based on data and research rather than intuition.
• Improved recruitment and talent acquisition : Enhancing hiring processes and improving the quality of hires through recruitment process data analysis.
• Better employee performance and productivity management: Identifying top performers and areas for improvement.
• Effective workforce planning: Forecasting future workforce needs and planning accordingly.
• Skills gap analysis : Identifying and addressing gaps in employee skills.
• Employee turnover prevention : Understanding and addressing the reasons behind high turnover and reducing flight risk .
• Identifying inefficiencies: Spotting and rectifying inefficiencies within the organization.
• Cost savings: Reducing unnecessary expenses through informed decision-making.
• Improving workplace safety: Enhancing safety measures based on data analysis.
• Uncovering trends: Gaining insights into workforce trends to inform HR strategy .

The 4 types of HR analytics explained

1. Descriptive analytics

How descriptive analytics works.

• Assessing behavior
• Comparing characteristics across time
• Spotting anomalies
• Identifying strengths and weaknesses

Advantages and disadvantages 

Descriptive analytics examples

• PTO : Using descriptive analytics, HR can analyze the average number of paid time off days that employees use in one year.
• Employee turnover : Descriptive analytics could be used to analyze employee turnover rates to compare the annual turnover between two teams or two departments. 

2. Diagnostic analytics

How diagnostic analytics works.

• Identifying the patterns and anomalies within the data that raise questions and need to be studied further.
• Discovering what factors could be contributing to the patterns and anomalies to identify the relevant data.
• Determining causal connections by analyzing the data with various methods.
• Data drilling: Taking information from a more general overview and providing a more granular view of the data.
• Data mining: Extracting patterns from data to help predict future events
• Probability theory: Quantifying uncertain measures of random events 
• Regression analysis: Determining which variables will impact an outcome
• Correlation analysis: Tests the relationships between variables
• Statistical analysis : Collecting and interpreting data to determine underlying patterns

Diagnostic analytics advantages and disadvantages

Diagnostic analytics examples

1. employee absenteeism, 2. employee engagement, 3. predictive analytics, how predictive analytics works, predictive analytics advantages and disadvantages.

Predictive analytics examples

1. recruitment, 2. retention, 4. prescriptive analytics.

• Machine learning
• Artificial intelligence
• Pattern recognition

How prescriptive analytics works

Prescriptive analytics advantages and disadvantages.

Prescriptive analytics examples

1. staffing, 2. attrition, to conclude, weekly update.

Stay up-to-date with the latest news, trends, and resources in HR

Andrea Boatman

Related articles.

15 Important Change Management Metrics To Track (In 2024)

11 Important Employee Relations Metrics To Track

Candidate Experience Metrics: How To Measure and Improve Candidate Experience

New articles.

360 Recruitment: Your Ultimate 2024 Guide

People and Culture vs. HR: What’s the Difference?

54 Employee Feedback Survey Questions To Ask (In 2024)

Subscribe to our weekly newsletter, are you ready for the future of hr.

Learn modern and relevant HR skills, online

How to be great at people analytics

A decade ago, someone touting the benefits of “people analytics” probably would have been met with blank stares. Was there value to be gleaned from HR data? Absolutely. But firms were thinking more narrowly about the potential—focusing on core HR systems and gathering straightforward information, such as snapshots of regional head counts or the year’s average performance evaluation rating, rather than using analytics capabilities to manage talent and make evidence-based people decisions.

Today, however, the majority of large organizations have people analytics teams, 1 Innovation generation: The big HR tech disconnect 2019/20 report , Thomsons Online Benefits, July 24, 2019, thomsons.com. 70 percent of company executives cite people analytics as a top priority, 2 “How people analytics can change an organization,” Knowledge@Wharton, May 23, 2019, knowledge.wharton.upenn.edu. and there’s little argument that people analytics is a discipline that’s here to stay. What’s striking, though, is the different ways that firms have approached building their people analytics functions. Team size, composition, and organization vary widely, and priorities for capability development and maturation differ significantly.

Most companies still face critical obstacles in the early stages of building their people analytics capabilities, preventing real progress. The majority of teams are still in the early stages of cleaning data and streamlining reporting. Interest in better data management and HR technologies has been intensive, but most companies would agree that they have a long way to go.

Leaders at many organizations acknowledge that what they call their “analytics” is really basic reporting with little lasting impact. For example, a majority of North American CEOs indicated in a poll that their organizations lack the ability to embed data analytics in day-to-day HR processes consistently and to use analytics’ predictive power to propel better decision making. 3 Based on responses of participants at a McKinsey roundtable of 45 chief human-resources officers in the autumn of 2016. Frank Bafaro, Diana Ellsworth, and Neel Gandhi, “ The CEO’s guide to competing through HR ,” McKinsey Quarterly , July 24, 2017. This challenge is compounded by the crowded and fragmented landscape of HR technology, which few organizations know how to navigate.

So, while the majority of people analytics teams are still taking baby steps, what does it mean to be great at people analytics? We spoke with 12 people analytics teams from some of the largest global organizations in various sectors—technology, financial services, healthcare, and consumer goods—to try to understand what teams are doing, the impact they are having, and how they are doing it.

Stairway to impact

It helps to think about the growth trajectory of a people analytics team as a stairway with five steps (Exhibit 1). The best teams don’t climb directly from one step to the next one; they are constantly iterating—retracing their steps and climbing the same stairs again—at every level of the journey to the top.

To move from the first step of the stairway (poor data) to the second step (good data), an organization must focus on building a foundation of high-quality data. This usually means that data needs to be extracted from the transactional systems where it is entered and then reshaped, cleaned, and re-coded into a more manageable and easier-to-understand structure that is aligned to the goals of the people analytics team. The more that analysts and data scientists need to clean and recode data to make it usable for even simple analysis, the less efficient the analytics team will be and the longer it will take to develop its skills and capabilities. This is arguably the most difficult step to get right. Significant resources, time, and investment are required to identify and manage core HR data systems, establish a common language and consistent data structure, and determine a basic set of guidelines for data collection, processing, and engineering. These are iterative processes, with no definitive solutions; rather, the processes and their outcomes change as the internal and external talent environments shift, systems are retired and renewed, and links are established among HR teams such as recruiting, training and development, and employee benefits.

As the operating environment changes at an increasingly rapid pace, both capabilities and the technology used to manage and transform data need to be increasingly flexible. In people analytics, as in many other tech-enabled fields, taking an agile approach is now a fundamental requirement. People analytics teams must work together with their enterprise-wide technology groups in a rapid and nimble way to institute new technology platforms, evolve existing infrastructure, and maintain consistent enterprise-wide standards.

Once a strong data foundation is in place, the people analytics team can climb to the third step, making the useful data accessible to the organization and experimenting with new technologies to analyze and disseminate the data. The sophistication that organizations are able to achieve at this step is variable. At the simplest end of the spectrum, teams might focus on automating and visualizing HR dashboards via standard business-intelligence platforms such as Tableau, in order to generate standard reports or respond to ad hoc requests. More advanced teams might prioritize custom builds and software development for self-serve applications, perhaps using their own front-end developers.

It’s evident from our interviews that organizations arrive in different ways at the ability to put data and actionable insights into the hands of decision makers. At several points, organizations must make decisions related to technologies and platforms—decisions such as whether to use homegrown talent or third-party vendors—and the answers vary by organization. As one would expect, the ability to attain advanced automation and self-serve capabilities depends greatly on the quality and accessibility of the underlying data.

Teams that mastered descriptive and automated reporting at step three are ready to climb to step four and build advanced-analytics capabilities. Data scientists, rather than business-information specialists, use programming languages like R, Python, and Julia to join disparate sources of data, build models to help understand complex phenomena, and provide actionable recommendations to leaders making complex and strategic business decisions.

We spoke to people analytics teams at a handful of organizations that are experimenting heavily at this level of the stairway and still have significant room to grow as their companies become open to new statistical tools, scale their data-science talent bench, and pursue a wide range of use cases. While some companies employ “broad-spectrum” data scientists who work cross-functionally to support a wide range of business needs, we found that the most advanced teams have created specific subspecialties in data science (for example, natural-language processing, network analytics, and quantitative psychometrics). These allow people analytics teams to increase their impact on their organizations by providing the advanced insights necessary to support strategic decision making on diverse and complex types of talent issues.

No people analytics team we interviewed has been able to take a full fifth step to reach the top level of the stairway: creating reliable, consistent, and valid predictive analytics. Reliable predictions will enable people analytics teams to analyze and explore practical options for management action. While some organizations have built fit-for-purpose predictive models—mostly for workforce planning—implementing predictive analytics in the context of employee selection, development, or engagement decisions requires a substantially scaled-up data-science operation, massive amounts of highly accurate data (“very big data”), cutting-edge algorithmic technology, and organizational comfort with how to address the impact on fairness and bias.

Beyond the required resources and the complexity of the analytics techniques, the infrastructure also poses a challenge to scalability and could require the use of cloud services. Most of the teams we spoke with are still working from on-premise technological infrastructures and show few signs of migrating their data and analytics capabilities to cloud services in the near future.

Ingredients for success

Our conversations with people analytics teams in leading organizations reveal a set of six best-in-class ingredients that have helped to propel the teams’ impact, success, and continued growth. These ingredients fall into three main categories: data and data management, analytics capabilities, and operating models. If we were to build a leading people analytics team from scratch, this is what we would strive for.

Data and data management

All great analytics teams are enabled by strong data standards, engineering, and management, and our interviews confirmed that this is no different in people analytics.

Significant and dedicated data-engineering resources. We found that the greatest team differentiator was the level of dedicated data-engineering resources available to it for propelling data creation and quality control. The leading teams have full ownership of their own data repositories, allowing them to rapidly test new ideas, iterate, and reduce dependencies on enterprise-level technology resources.

An added benefit of dedicated data-engineering resources is that they enable strategic alignment. Data engineers who are steeped in the strategic context of their organization’s people analytics teams are able to design the data foundation and analytics solutions more thoughtfully and deliberately from the beginning.

Breadth and depth of data sources. Leading teams have invested heavily in a strong HR-data foundation but also have advanced ways of going beyond the core HR systems to use several additional internal sources of data. The most straightforward way might be seamlessly linking the HR data with finance data, though data priorities will differ depending on organizational context. A few teams have begun to step beyond relational databases to build graph databases 4 A type of NoSQL database, graph databases are able to model relationships within data in a powerful and flexible manner. For more, see Antonio Castro, Jorge Machado, Matthias Roggendorf, and Henning Soller, “ How to build a data architecture to drive innovation—today and tomorrow ,” June 3, 2020. for advanced network analytics. In addition, leading teams have a robust and flexible survey strategy for monitoring employee sentiment. They are also able to integrate their survey data with multiple other data sources to create multidimensional quantitative and psychometric models that help explain employee engagement trends and dynamics.

While it is common for people analytics teams to feel constrained by a lack of easily available data, leading teams are more creative with data, acquiring new sources or combining existing ones in new ways to attack the problem at hand. For example, time-sheet data could be transformed and loaded into a graph database and linked by activity or project codes to allow better analysis of teamwork and collaboration.

Analytics capabilities

Advanced people analytics projects can require both deep technical knowledge and the ability to integrate and translate across a wide array of expertise and input. The best teams are building their talent bench with breadth and depth.

Robust data-science function. As we expected, all the leading people analytics teams we interviewed have invested heavily in acquiring data-science talent, though their approaches differ. Some teams focus on hiring “all-around athletes,” while others prioritize specialized backgrounds such as quantitative psychometrics or natural-language processing. Leading teams have sizable data-science “pods” that span a wide range of advanced analytical methodologies, programming languages, and academic backgrounds. The best teams hire and develop specialists in specific disciplines of data science but nevertheless expect all of these individuals to operate in a nimble, cross-functional way in order to meet evolving needs.

Strong translation capability. Leading teams also complement their high-caliber technical talent with skilled “translators”: specialized “integrators,” who bridge the gap between business leaders and technical experts. They translate strategic challenges into analytic questions and use evidence-based practice to interpret insights derived from the analytics, engage stakeholders, and ultimately propel business changes. Translators often serve as an entry point to the people analytics team, helping to raise awareness of the team in the organization and build the team’s credibility. Some of the leading people analytics teams have built benches of internal consultants to partner directly with individual businesses on their specific problems.

Operating models

In a fast-developing field, people analytics teams need to deliver impact across the organization and stay ahead of the curve to maintain that impact into the future. The best teams align themselves well against organizational priorities while maintaining space for open experimentation and innovation.

Innovation as the norm. Members of leading teams are explicitly expected to explore and innovate beyond their day-to-day fulfillment of the needs of their clients. Some companies have rules of thumb for the percentage of time that teams spend on exploration as opposed to core foundational work. These expectations allow teams to fully experiment and build out proofs of concept.

This process can take a variety of forms, but the important distinction is that the areas of innovation need not directly support an existing business priority or client need; they might be purely exploratory. For example, some data scientists relish the extra time to play around in a sandbox and learn how analytic tools and services work in the cloud. Others might want to explore creative new ways to visualize data in order to equip business leaders with helpful insights. The goal is to ensure that all team members are constantly forming new ideas and looking for new ways to meet the analytic needs of the organization and thereby help it achieve its objectives.

Clear alignment with clients and organizational use cases. People analytics teams take different approaches to organizing themselves and aligning with different clients. What is consistent, however, is the presence of a mechanism for attaining an in-depth understanding of enterprise-wide priorities as well as the specific needs of individual clients. This mechanism creates feedback loops that enable continuous learning and iterative development, and it ensures that people analytics teams are working on the most pressing and high-impact topics.

A culture of trust, empowerment, and ownership is the critical foundation for ensuring that a people analytics team is aligned with its clients as well as the enterprise. People analytics teams routinely deal with urgent (and often ambiguous) client needs and questions, highly sensitive data, and challenges to extrapolating meaningful and actionable insights that will guide business decisions. The bar to entry for the best teams is high: members must own their work from end to end and be empowered to define the constraints of any analysis, protect privacy as well as fairness and equity, flag issues that arise, and use their own judgment to derive insights. Being reactive and incremental is not enough in human resources, where priorities change and the top ones require immediate attention.

Over time, as organizations become increasingly dependent on the quality of their insights, the best people analytics teams play a stronger role in shaping the HR agenda, influencing how the organization manages its talent at both a policy and a process level.

The pulse survey

The COVID-19 crisis provided a natural experiment for one large, global organization with a strong people analytics team to use the ingredients outlined in the previous section by rapidly creating a homegrown weekly pulse survey to track the opinions and feelings of tens of thousands of employees around the globe. This capability enabled the organization to better understand the best ways to support employees in a challenging time and a fully remote work environment.

Setting up the pulse survey required intensive collaboration between diverse, highly skilled individuals already embedded in the organization’s people analytics team as well as rapid and close collaboration with the leadership of the organization. Translators navigated the need to craft questions that engaged employees, gathered high-quality data to feed the analytic models, and communicated insights back to leaders who had urgent decisions to make about how to best support their workforce in an external environment that was highly unpredictable and changing week by week.

To speed the time to insights, data engineers established an automated and continuous link among weekly survey-response data, core HR data systems, and a broader set of additional data sources, including data sets that data engineers had developed and customized for this purpose. This process cleaned, tested, and prepared the data for analysis. In addition to rapidly providing analysts with weekly data to examine and synthesize, it fed these data to a prototype self-service reporting tool, which gave leaders the ability to directly investigate aggregated pulse data within six hours of the survey’s close.

The customized data sets supported both exploratory and targeted analyses and helped generate actionable insights for the leaders. Analyses were designed to build on the organization’s current understanding of the health of its employees, marrying new and existing information to yield new insights that guided various efforts. For example, specialists in natural language processing used structural topic modeling to identify and quantify topics in the free-text comments that employees submitted as part of the survey each week. Sentiment analysis was used to understand the emotion behind each topic. These results were then married to the demographic information prepared by data analysts, allowing managers, leaders, and other decision makers to understand how the conversations and associated feelings varied by subpopulation, such as parents and less tenured employees. The combination of data sources and analytic approaches ultimately revealed population-specific needs, which allowed the organization to target specific groups and tailor the type of support it offered to maximize impact.

Exhibit 2 is a view of the major topics generated from the free text of the employees who responded to the pulse surveys and how their emphasis on these topics changed over the course of two months of the crisis. At the beginning, employees were thankful for the health of their families and peers and had generic concerns about the developing situation, but as the crisis evolved, their thoughts crystallized into the more particular concerns of isolation, remote work, childcare, and work-life balance.

The ability to rapidly develop this capability, turn around a wide range of sophisticated analytics within 24 hours after the survey closed, and repeat the survey weekly did not come easily to the organization or the people analytics team. The capabilities required to pull it off were tightly rooted in the data, analytics, and operating-model ingredients that we have identified as the hallmarks of great people analytics teams.

Despite the vast differences that exist among organizations’ data quality, integration, and infrastructure, we all certainly have a lot to learn from each other. Answering the following questions will be helpful to leaders who want to identify where their organization’s people analytics is now and where they would like them to be:

• Where is the organization on the people analytics stairway? Where does it aspire to be in the next year, three years, and five years?
• How does the organizational context influence the mandate of the people analytics team?
• What ingredients does the organization possess today, and which does it need to build?
• How should the organization determine its priorities in building people analytics capabilities? For example, should it build to support certain specific internal use cases, or should it build a broad bench of capabilities to support an unpredictable or rapidly changing internal environment?
• If the organization had to get one thing right over the next 12 months, what would it be? What would get in the way of its getting there?

While no single model is the “correct” one for developing the capabilities of a people analytics team, leading teams seem to have a set of ingredients in common. While the past decade has brought about real change, even the best teams—those that iterate at each step of the stairway and learn as they ascend—have barely scratched the surface of what’s possible with people analytics.

Elizabeth Ledet is a partner in McKinsey’s Atlanta office; Keith McNulty is a director, people analytics and measurement, in the London office; Daniel Morales is a director of analytics in the Washington, DC, office; and Marissa Shandell is an alumna of the New York office.

Explore a career with us

Related articles.

A transformation of the learning function: Why it should learn new ways

What 800 executives envision for the postpandemic workforce

The boss factor: Making the world a better place through workplace relationships

Towards a process-oriented understanding of HR analytics: implementation and application

• Original Paper
• Open access
• Published: 18 August 2022
• Volume 17 , pages 2077–2108, ( 2023 )

Cite this article

You have full access to this open access article

• Felix Wirges   ORCID: orcid.org/0000-0001-9939-6444 1 &
• Anne-Katrin Neyer 1  

10k Accesses

18 Citations

Explore all metrics

Firms have recognized the opportunities presented by HR analytics; however, it is challenging for HR to convert their available data (sources) into meaningful strategical value. Moreover, research on the implementation and application of HR analytics is still in its infancy. Drawing on the socio-technical system perspective, we examine the implementation and application of HR analytics in firms. Based on a qualitative study with 17 HR analytics experts, we find that a shift to a more process-oriented perspective on HR analytics is needed. More precisely, besides the requirements for the analysis of data, the actual roles in the process of implementing and applying HR analytics need to be defined. In particular, this implies the interaction between the specialist department, the HR business partner and the HR analytics function. From a managerial perspective, we propose a process model for the future implementation and application of HR analytics.

Similar content being viewed by others

The Praxis of HR Analytics

HR analytics between ambition and reality

HR Analytics: Serbian Perspective

Avoid common mistakes on your manuscript.

1 Introduction

Data collection and analysis is an essential part of business process decision-making for a variety of organizations (Côrte-Real et al. 2017 ; George et al. 2014 ). Data-based decision-making takes place in almost every area of an organization; marketing, sales, production and finance are examples of application areas where it is common to make decisions based on analysis and reporting. This is not particularly surprising, as analytics-based decision-making can automate processes and make them more efficient in almost every business area where data can be gathered (Acito and Khatri 2014 ; Earley 2015 ; Ghasemaghaei 2018 ).

However, compared to the extensive use of data in other application areas of the company, its use is rather rare in HR (Tursunbayeva et al. 2018 ). To help to deal with this endeavor, in recent years, the term "HR analytics" has become popular and has increasingly been used in science and practice, often promising nothing less than the revolving of the HR management (Marler and Boudreau 2017 ; van den Heuvel and Bondarouk 2017 ; Huselid 2018 ; McIver et al. 2018 ; Tursunbayeva et al. 2018 ; Greasley and Thomas 2020 ; McCartney and Fu 2022 ). Falletta and Combs ( 2021 ) note that while the amount of data and technology has increased significantly, the use of data to explain organizational issues is not in itself a novel tool. A variety of approaches exist, focusing on a more evidence-based approach to HR management (Lawler and Boudreau 2015 ). This emphasizes that the design of the HR function is per se predestinated for the use of data-based decisions:

First, HR is a business area that, in principle, could generate a large amount of data. Through the ongoing digitization of work processes, the use of mobile devices, the wearing of wearables or the use of company apps, employees generate different types of data. These include e.g. information about locations, communication, personal well-being and many other factors, which are of relevance for HR (Cascio and Montealegre 2016 ).

Second, due to the growing use of information and communication technologies in HR, the term eHRM has become increasingly popular (van den Heuvel and Bondarouk 2017 ). More recently, the use of HR analytics has taken the topic to a new level: The previous use of technology in HR mostly concentrated on operative support or descriptive reporting of key information, such as sick days of the workforce or employee turnover. Through the use of HR analytics, connections and conclusions can now be drawn (a) in each functional area of HR, but also (b) with data from other application areas. Predictive analyses can then be made from these results. However, if we look at the practical application of HR analytics, especially with a focus on the use of predictive analytics, we see a sobering picture. Falletta ( 2014 ) shows in a sample of 220 firms that only 15% place a strategic focus on HR analytics and that, as a rule, they do not carry out any predictive analyses, but only focus on reporting. A survey by Lawler and Boudreau ( 2015 ) published two years later provides similar results. Levenson and Fink ( 2017 ) noted that the term HR analytics has become a catch-all term to describe any handling of data and metrics in HR: “It has come to include anything numerical about talent and HR work. Examples include simple data reports, analyzing data integrated from multiple systems (e.g. performance and compensation), dashboards, making data available “on demand,” and true talent or “predictive” analytics” (Levenson and Fink 2017 :146). More recent definitions of HR analytics emphasize shifting the focus towards a process perspective (Mclver et al. 2018 ): HR analytics is not only understood as a tool where statistical methods are applied and the focus is on key figures, but as a systematic approach (Falletta and Combs 2021 ). Most recently, Falletta and Combs ( 2021 ) define HR analytics as follows: “HR analytics is a proactive and systematic process for ethically gathering, analyzing, communicating and using evidence-based HR research and analytical insights to help organizations achieve their strategic objectives” (Falletta and Combs 2021 : 3). The two authors highlight that in the application of HR analytics so far, there is too little recognition “of the role of broader HR research and experimentation as part of an overarching HR analytics agenda (i.e. internal HR research or partnership research in the context of social, behavioral and organizational sciences)” (Falletta and Combs 2021 : 54). This goes in line with the evidence-based review by Marler and Boudreau ( 2017 ), which uses an integrative synthesis of published peer-reviewed literature. Their findings emphasize that HR could soon be technically left behind and thus, hint at an issue that has already been discussed for some years: HR must create technological change in order to continue to play an equal role in the company in the future (Shrivastava and Shaw 2003 ; Snell et al. 2002 ; Ulrich 1997 ).

Third, in the wake of the COVID-19 pandemic, it became clear that metrics-based information can be of tremendous importance. Companies were confronted overnight with unprecedented challenges in managing employees in the workplace. The HR function in particular had to ensure within a very short time that the new requirements for remote work, digital collaboration and leadership in teams, and (mental) health issues were met (Kniffin et al. 2020 ). Along the way, previous tasks such as recruiting or HR development had to be transformed into digital solutions. Companies, therefore, raised information about remote work, employee engagement, and well-being to gain a clear picture of the respective needs of employees (Belizón and Kieran 2021 ). The situation created by the pandemic highlights the importance of HR analytics. It essentially offers the possibility of many new types of data sources that can specifically promote the quality of HR analytics (Bryce et al. 2022 ).

Having said this, the pressing issue thus is: why do only a few organizations rely on the (advanced) use of HR analytics, although the circumstances (data generation, software applications, etc.) and the reason (strengthening the strategic role of HR) (Bassi et al. 2012 ) seem to be predestined for an application? Building on this question, we conducted an initial descriptive survey with HR employees and managers to determine the status quo of the implementation of data-based analytics (Wirges et al. 2020 ) (see Table 1 ).

The results of the study showed that working with data plays an important role in HR management. 66% of the interviewees said that data evaluation in HR is included in strategic decision-making. However, it should be noted that these decisions are predominantly based on classic HR controlling (Excel, KPI etc.) or descriptive analyses (the analysis of data related to the past). For example, 96% of the interviewees stated that they use the data obtained for HR controlling. In this study, HR controlling was broadly defined as the simple collection and reporting of individually defined key performance indicators. The use for descriptive analyses is already significantly lower at 32%. Only 5% carry out predictive analyses with the help of the data.

The results of this study reflect a sobering picture regarding the application of HR analytics. This underlines the current state in the literature that the knowledge about the implementation and application of HR analytics is fraught with many challenges and difficulties. For a better understanding of these challenges and to present them in a holistic picture, the aim of this paper is to dive deeper into the implementation and application of HR analytics. To do so, we conducted a qualitative study with HR analytics experts. By applying a socio-technical approach as a theoretical lens, we aim to answer the following research questions: How is the implementation and application of HR analytics shaping up in firms? What challenges and barriers do firms face on their journey towards HR analytics?

2 Theoretical background

To answer our research questions, we apply a socio-technical approach. Socio-technical system theory assumes that new systems can only be successful if both the technical and the social system are considered, analyzed and brought into harmony with each other on an equal footing (Cherns 1976 ). The social system describes the people in an organization and focuses on their needs, relationships and qualifications within the organization. The technical system, on the other hand, often describes novel technological artifacts used to accomplish tasks (Jaffee 2001 ; Mumford 2003 ). The socio-technical paradigm is a holistic view that examines the relationships between the social and technical levels of any system (Trist and Bamforth 1951 ; Coakes 2002 ). Socio-technical design emphasizes the need for an optimal match between the technical and social aspects in terms of the relationship between jobs and people's needs and expectations (Biazzo 2002 ). As discussed earlier, understanding HR analytics from a systematic process perspective has gained importance (e.g. Falletta and Combs 2021 ). In the literature to date, there are initial approaches to linking the socio-technical approach in the context of HR analytics (Belizón and Kieran 2021 ). Nevertheless, this can be classified as rather novel. Thereby, the goal of HR analytics is to enable organizations to achieve their strategic objectives. Since HR analytics comprises more than the introduction of software for personal data analysis, it requires a more holistic approach rather than a traditional IT project approach. Often, projects of this kind fail not because of the technology, but because of a lack of consideration of the mutual interactions of the social and technical system. Maucher et al. ( 2002 ) show that soft factors such as communication, cooperation and the inclusion of informal structures can contribute significantly to the successful implementation and application of IT projects.

Following the socio-technical system theory proposal we analyze the extent to which the technological side, i.e. the artifact needed for the data analysis and the social side, i.e. the organizational structures of a company with a large number of different stakeholders, need to be reflected by HR analytics. We conducted a literature review and classified the challenges we found for the implementation and application of HR analytics within the sociotechnical perspective into the two categories of social and technological. This involved searching for peer-reviewed journal articles that address challenges in the implementation and application of HR analytics. For this purpose, the following search terms were used to identify relevant articles: "HR analytics"; "People analytics"; "Human resource analytics"; "Workforce analytics"; "Data-driven HR" in combination with "challenges"; "difficulties"; and "barriers". The focus in the selection of the respective journal articles was on the thematization of concrete examples of implementation and application difficulties. In doing so, we were able to identify four core areas (see Fig.  1 ) that influence the implementation and application of HR analytics from a systematic process perspective.

Core areas influencing HR analytics

The first of the areas of the technological system we examine involves the data, which represent the elementary cornerstones for carrying out the analyses (Douthitt and Mondore 2014 ; Pape 2016 ). One is the quantitative aspect of analyzing whether or not the necessary HR databases with the associated data sources exist for the use of HR analytics. On the other hand, it must be seen whether the existing data meet the qualitative requirements in order to be able to carry out valid analyses (Jeske and Calvard 2020 ; King 2016 ; Minbaeva 2018 ; Pape 2016 ). Previous research shows the need for the integration of additional data from different areas of the company into the analysis (Marler and Boudreau 2017 ; McIver et al. 2018 ; Rasmussen and Ulrich 2015 ). To do so, interface compatibility plays an important role (Andersen 2017 ; Boudreau and Cascio 2017 ; Douthitt and Mondore 2014 ; Levenson and Fink 2017 ). Angrave et al. ( 2016 ) emphasize this by warning against data silos formed within the individual departments leading to a lack of data exchange. Indeed, the lack of available data has been identified as one of the main obstacles to the successful implementation of analytics, especially in small and medium-sized enterprises (Pape 2016 ).

The second area at the technological level comprises the technology itself i.e. which software and hardware solutions are available to users (Angrave et. al 2016 ; Aral et al. 2012 ; Boudreau and Cascio 2017 ; Douthitt and Mondore 2014 ). A lot of firms still stick to simple spreadsheet programs such as Excel for data analysis (van den Hauvel and Bondarouk 2017 ), even though in recent years the number of other tool providers, offering a wider range of functions for data analysis, has increased. However, so far, the available tools for predictive and prescriptive HR analytics are developed by and aimed at people with analytical skills, not HR business partners (Fernandez and Gallardo-Gallardo 2021 ). This is also shown by the results of our descriptive study, in which we found that the available software solutions are perceived to be too complex. More precisely, software solutions for the application of HR analytics are not tailored to the competencies of the users and, thus, will need to be much more user-friendly (Marler and Boudreau 2017 ). The application of the technology, i.e., software solutions, therefore depends strongly on the respective competencies of the HR business partners or the users of HR analytics in a company.

This assumption simultaneously emphasizes the interconnection of the technological with the social system. A strong social system lays the foundations for the implementation and application of HR analytics. Within the social system, we first focus on the HR business partner as a user of HR analytics (Bassi 2011 ; Mondare et al. 2011 ; Rasmussen and Ulrich 2015 ; Angrave et al. 2016 ). It is generally agreed that one of the main reasons for the low use of HR analytics is the lack of analytical skills (Angrave et al. 2016 ; Marler and Boudreau 2017 ). However, these analytical skills are an elementary prerequisite for performing HR analytics (Andersen 2017 ; Douthitt and Carson 2011 ; Huselid 2018 ; Kryscynski et al. 2018 , Marler and Boudreau 2017 , Minbaeva 2018 , van der Togt and Rasmussen 2017 ). Prior studies have analyzed that individuals working in HR are not primarily interested in operating with key figures, statistical methods or data analysis (Rasmussen and Ulrich 2015 ). Fernandez and Gallardo-Gallardo ( 2021 ) emphasize that the analytical skills needed to apply HR analytics will increase in the future. Thus, a wider range of basic statistical methods in the individual maturity levels of data analysis (reporting, descriptive, predictive, prescriptive) and analytical competencies in data collection and data management are considered important to HR analytics (Levenson 2011 ). Because of the advancing increase in artificial intelligence and its methods such as machine learning, we assume that HR business partners will be required to continuously learn and extend their knowledge (McIver et al. 2018 ). The personnel development measures required for this in turn have a positive effect on the attitude toward HR analytics and increase the individual's self-efficacy (Vargas et al. 2018 ).

The intra-organizational context is also relevant in explaining the influence of the social system on the implementation and application of HR analytics. First, the focus is on the different stakeholders involved in the process (Coco et al. 2011 ; Giuffrida 2014 ; Levenson 2011 ; Rasmussen and Ulrich 2015 ). These can be divided into HR business partners, management, employees and analysis teams (Huselid 2018 ; Peeters Paauwe and van de Voorde 2020 ). In previous research, there are few findings about which stakeholders are involved in the process of HR analytics (Coco et al. 2011 ). However, we argue that our analysis shows they do not look at the respective roles and relationships in the implementation and application of HR analytics in an organization. Second, there are two perspectives that explain how HR analytics should be embedded within the organization, i.e. outsourcing and integration (van den Heuvel and Bondarouk 2017 ). The two different perspectives can be characterized as follows: In outsourcing, the analyses are carried out by experts in the analysis area and not by the actual HR business partners. The HR analytics function operates independently alongside the traditional HR function (Fernandez and Gallardo-Gallardo 2021 ; Rasmussen and Ulrich 2015 ). In the case of integration, an attempt is made to strengthen the competencies of the HR business partner and to perform the analyses within the HR function with the help of HR analytics. Outsourcing HR analytics from the HR function is supposed to align HR analytics more strategically (Rasmussen and Ulrich 2015 ). In contrast, its integration into the HR function (Angrave et al. 2016 ; Bassi 2011 ; Falletta and Combs 2021 ), will enable HR to strengthen its own strategic role. Additionally, it is argued that the expertise of HR business partners in the respective functional areas is needed. If they are not directly integrated into the process, the usefulness of the analysis of HR-specific issues can only be assessed to a limited extent (Andersen 2017 ).

Having presented our framework, we follow Greasley and Thomas ( 2020 ) call for further empirical analysis of analytics projects to conduct research in HR analytics with a focus “on the process of development rather than its outcomes” (Greasley and Thomas 2020 : 506).

3 Methodological approach

In order to gain deeper insights into the implementation and application of HR analytics, we conducted a qualitative study with 17 HR analytics experts from the DACH region. Our study aims to understand the process of implementing and using HR analytics in more detail Therefore, we use a qualitative research approach, which is particularly suitable for investigating topics that have been little empirically researched so far and require a deeper insight into situational conditions. Moreover, qualitative research also lends itself specifically to the representation of organizational processes, as one can derive important information about social interactions and causal relationships from the depth and variety of data obtained (Graebner et al. 2012 ).

Hyde ( 2000 ) notes that the information content in qualitative research is based on the depth of the interviews and even the knowledge of one person, if the rules of qualitative social research are followed, can provide insightful knowledge about complex issues (Hyde 2000 ). Thus, the identification of the experts was the first crucial step in our study. A targeted search was conducted via job-related social networks for job titles that included the competence profile HR analytics, people analytics, workforce analytics or HR executives who mentioned working with data in the HR management in their competence profile. To make sure that only HR analytics experts take part in the qualitative study, the selection of participants study was based on the following criteria:

The interviewees explicitly deal with the topic of HR analytics in their company and already have experience in its implementation and application.

The respective maturity level of the application of HR analytics (reporting, descriptive or predictive) in the company played a subordinate role in order to gather as much experience as possible.

The extent of the professional experience with HR analytics of the interviewees also played a minor role, as many firms are only in the early stages of HR analytics.

Each interview was conducted using a semi-standardized guide (see " Appendix "). The basis for this was the previously deductively formed socio-technical framework with the four categories of data, technologies, personnel deployment and organization. The interviews were then transcribed and analyzed using the atlas.ti software. A total of 200 pages of transcribed interview data were collected. The interviews lasted an average of 45 min. The interviews were conducted and transcribed in German. The results of the qualitative content analysis were translated into English.

Figure  2 illustrates our methodological approach. We proceeded in three steps, which are explained in the following. At the beginning of the interview, it was important to capture the interviewees’ understanding of HR analytics, given that there is no uniform definition of HR analytics. Therefore, the interview partners were asked about their task profile and the current state of application of data-driven decisions in their respective firms. This enabled a classification of the status quo in the subsequent analysis of the interviews (see 1st methodological step). Thereupon, specific questions were asked about the current application of HR analytics. In qualitative research, two general approaches can be distinguished: on the one hand, the frequently used inductive approach, which creates a generalization on the basis of specific observations. On the other hand, there is the deductive approach, which tries to transfer generalizations to a specific case (Hyde 2000 ). We initially have chosen a deductive approach according to Mayring ( 2014 ) and a category system was developed on the basis of the factors that have already been derived in our framework as influencing the implementation and application of HR analytics. In a first step, definitions for the individual categories were assigned and suitable anchor examples and coding rules were determined (see Table 2 ).

Three-step methodological approach

The individual interviews were then analyzed and individual text passages could be assigned to the respective categories on the basis of the predefined rules. Based on this, we structured our analysis along the four categories to identify the problems and challenges mentioned by the interviewees. This process allowed us to identify specific aspects within the theoretically developed framework of existing requirements for the implementation and application. In addition to the four deductive main categories of data, technology, HR business partner and organization, further subcategories were inductively formed in the next step. In contrast to deductive coding, inductive coding is based on the principle of open coding. This means that the respective statements of the interviewees are openly coded in the first step and that more aggregated categories emerge from the raw data through repeated examination and comparisons. Ultimately, this allowed for the formation of further subcategories: database, application area, tools for analysis, tools for the provision of the results, current role model, future role model, management support and added value (see 2nd methodological step). Building on the category system, we then specifically searched for patterns of interaction and process within these categories in order to focus on the systematic process perspective (see 3rd methodological step). We proceeded as follows: The interview material was examined one more time for statements describing interaction processes between individual actors in the context of HR analytics. In coding, we defined interaction as the mutual influence between actors. The focus was on the communication and action processes described by the respondents. For example, the following statement describes the provision of analysis results from the HR analytics function to HR business partners.

So the skillset definitely has to grow and we are also in the process of taking the first step and want to make that available as a service, where now the HR business partner, for example, only has read access (Interview 3).

Upon further coding, additional statements could be found that confirmed this process of providing analysis results. In this way, consistent patterns of interaction between different stakeholders could be derived from the individual statements of the interviewees and presented aggregated in the form of a model. This model we derived represents the status quo of HR analytics from a process-oriented perspective.

4.1 Current state of application of HR analytics

We begin the presentation of our results with a brief description of the current state of application of HR analytics in the respective firms. In doing so, we highlight the general understanding of HR analytics and address the changes within the processes of HR. In the next step, we present the aggregated findings from the respective categories, which are based on our framework data, technology, HR business partner and organization. Table 3 provides an overview of the 17 interviewees.

Our analysis of the current state of application resulted in the majority of respondents still being in the early stages of using HR analytics. Many of the firms are currently in a start-up phase, which is characterized by conducting more in-depth descriptive analyses and answering isolated predictive questions. The analysis of how the interviewees define the term HR analytics is characterized by a uniform understanding. HR analytics is a strategic tool that offers the possibility to steer and influence actions and decisions in the HR context on the basis of analyses. The interviewees emphasized that one does not rely exclusively on these analysis results, but that they should be considered as decision-supporting. Rather, the analyses with the help of HR analytics are intended to stimulate the HR business partners to take a deeper look at topics, as one of the interviewees emphasized:

So, there's also a lot of show and tell in the collaboration with HR staff, rather than them sitting down, doing some calculations themselves and thinking afterwards, okay, that'll get us XY. Yes, it's a people business, and it's also a very emotion- and perception-driven business. And then you can use good and relevant analytics to consistently influence people's perceptions and actions (Interview 9).

4.2 Delimitation HR controlling

The goal of HR analytics is to use the methods applied to HR data to check the effectiveness of HR measures and ultimately identify levers that can contribute to improving processes within HR and also the entire company. In particular, the interviewees underlined a demarcation from classic HR controlling. While HR controlling mainly summarizes key figures on historical data and provides relevant groups with information, analyses with the help of HR analytics aim at looking at these key figures in more detail, deepen them and apply them in future-oriented decision-making processes. Our findings underlined that this process is not solely based on predictive HR analytics. In contrast, the interviewees highlighted the potential of descriptive HR analyses, while at the same time emphasizing that it is difficult to manage the next step towards predictive analyses. Additionally, it is found that it can be difficult to draw a line between HR analytics and HR controlling given that the boundaries are sometimes blurred. For illustration, one of our interview partners emphasized that in some cases HR analytics is understood as the automation of reporting processes.

I think it's a big problem. Because then you are actually leading the absurdity of what this three-pillar model is supposed to achieve. And everything that people analytics is supposed to stand for, namely to really bring a benefit to the business and not just to run some purely administrative evaluations somewhere. Yes, definitely. Unfortunately, that's what many organizations have done (Interview 14).

Based on the initial maturity level in which the interviewees find themselves with regard to the implementation of HR analytics in their firms, it can generally be stated that HR analytics is divided into standardized advanced reporting, which has a descriptive character, and project-related questions, which go deeper in the type of analysis. Table 4 presents the results of the qualitative content analysis in condensed form. These results are based on the coding of the interviews and are explained in more detail below.

4.3 Technological system: data

4.3.1 database.

Our data showed that the data basis is an important challenge for the use of HR analytics. However, it should be noted that firms with a well-functioning HR controlling system have a solid database. These are standard data from the HR management, such as fluctuation figures, sick leave, master data, salaries, etc. One interviewee pointed out that especially firms with little or weak digitization of processes have to struggle with data quality problems. It was pointed out that one of the most important first steps in the implementation of HR analytics should be the creation of an all-encompassing and aggregated HR system, otherwise one is busy with manual data cleansing, especially in the beginning.

If you have an outdated HR system or an old SAP HRM system, you are extremely immobile in the use of the data because you can only get it out with difficulty or not necessarily in the format you would like to have and then you are already back in this operations trap. That is, you come up with a great dashboard, great use case and build it within a week and then you have to download a report every week for the rest of your life, pull it in, maybe clean it up a bit. And these are exactly the pitfalls that you run into (Interview 9).

Our findings showed that firms have a hard time with data integration in particular because HR does not store its data centrally in an enterprise data warehouse, as many firms do out of caution. Consequently, business intelligence topics have also not found access in HR for a long time, which results in a poor-quality database.

4.3.2 Application areas

Our analysis reveals that even though HR analytics is applied in different HR functions, all interviewees emphasized that recruiting can be identified as the most effective area. On the one hand, this is due to the large number of data records generated by applications, which makes it possible to develop valid prediction models compared to other application areas. On the other hand, recruiting is also seen as having the greatest potential for highlighting the added value of HR analytics given that it is a major cost factor for many firms. However, it has also been pointed out that rather “new” topics are suitable for data-based analyses, as the attitude toward these newer topics has not yet solidified in the minds of those involved. One interviewee highlighted that this is specifically the case for diversity, i.e. issues such as equal pay, women's quota, and severely disabled quota, as these are high visibility issues where companies are generally more open to learning more.

Because these are fields that have only been established in this way for a few years. And that's where the knowledge has to be built up. So, there is just less gut feeling or felt gut feeling and therefore more room for such analyses (Interview 1).

When using HR analytics, the potential benefits of the individual questions in the application areas should be considered. For instance, one of the interviewees emphasized that the economic benefit should be kept in mind during the analysis.

I [think we] consider far too little in HR analytics or people analytics, that we align ourselves with the business problems (Interview 6).

For a small company with a large turnover, an analysis of the reasons for turnover can create a relatively large added value, whereas on the other hand a large production group, with an identical turnover has a saving that is not significant, but a potential analysis to improve the ergonomic working conditions on the assembly line can increase productivity.

Whereas the orientation of analyses in the business context is crucial, another interviewee also underlined that especially in the beginning the process of quantifying all possible processes in HR management can contribute to developing a feeling for dealing with numbers and can create an orientation to include analyses in the decision-making process.

So everything that is really purely statistical figures first of all in the personnel area. That's good. To be honest, I think it's also important because it helps you to awaken a bit of an affinity for numbers or a feeling for them. But anything that goes beyond a standard evaluation, I would always tie to a concrete business case (Interview 14).

One interviewee also noted that " you don't want everything you can " (Interview 12). He emphasized that one must consciously look at whether the analyses make sense for the respective industry and really lead to an increase in effectiveness.

4.4 Technological system: technology

4.4.1 tools for analysis.

Besides the data, another technological aspect is the technology itself, i.e. the analysis software used. Here, a largely homogeneous picture emerged among the interviewees. Tools such as Tableau or PowerBI were used for the actual analysis. In rare cases, additional work was done with R or Python. This is mainly the case in firms, which already carry out more advanced analyses. Two firms worked with external analysis tools. One of these is an external software manufacturer that offers a stand-alone HR analytics tool and the other is an integrated analysis function of the HRIS.

Other interviewees were rather critical of the use of external tools, as there is no direct insight into the analysis methods. As an example, tools were cited that offer e.g. speech analyses of interviews that are supposed to predict suitability without providing valid evidence for this. The use of such tools has a counterproductive effect and stirs up fears. Another aspect that has been criticized about external tools is their limited flexibility. An individual adaptation to the circumstances of a company is only possible to a limited extent. Predefined standard use cases may be applicable, but they often cannot be specifically adapted to the particularities of the company structure. This is particularly important in HR management, as it is characterized by a high degree of variance, as the following interviewee noted:

There is a lot of variance in what an organizational structure looks like. Do you have double tops or just single tops, do you have a pyramid or a cell structure? These are all issues that have an extreme influence on the data model that you have to import into such a system. And for the fact that I pay relatively a lot of money for relatively simple dashboards that come out of it, I think that's pretty meager (Interview 9).

4.4.2 Tools for the provision of the results

With regard to the question of the technology to be used, a differentiation must be made between the process of the actual analysis and the provision of the results of this analysis. This is strongly related to the understanding of roles in the HR analytics process. This aspect will be discussed in more detail in the course of the results. The presentation of analysis results is provided to the respective addressees via a tool such as Tableau. Here, the interviewees saw the focus above all on ease of use. The simplicity and instinctive approach were emphasized. One of the interviewees pointed out that it is precisely this simplicity that also empowers and motivates people to work with metrics. Furthermore, the flexibility for visual representations was emphasized, which is particularly important for HR management in order to provide HR business partners with a more comprehensible approach to the subject matter.

When choosing the respective tools, it is necessary to define in advance exactly who plays which role in the HR analytics process, as the following interviewee pointed out:

So, we say, you can introduce the best tool if just this, the use of the tool is not clear. So, if the end-user is not clear where he is going to use this (Interview 3).

4.5 Social system: HR business partner

Another aspect in the investigation of our data was the HR business partner. Here we were able to identify the current role of the HR business partner and the problems associated with it. We were also able to identify the extent to which the understanding of the role must change in the future for the effective application of HR analytics.

4.5.1 Current role model

If we first look at the statements regarding the competencies of the HR business partners, it became clear that the analytical competencies of the HR business partners were predominantly assessed as poor to barely present. In the eyes of the interviewees, the HR business partner is not the one who carries out in-depth analyses. Our analysis showed that there are three main reasons for this.

First, one of the reasons lies in the nature of HR. The background of working in HR is often different from working with data and key figures, so many of the current HR business partners have avoided basic statistical subjects already in their studies and thus have not developed a connection to data-based analyses in the course of their professional life. Secondly, there is a lack of understanding and rejection of working with data in HR. The third aspect lies in the number of operational tasks and lack of time highlighted by the interviewees. Even if HR business partners are willing to build competencies in the area, this often does not happen due to time constraints.

So it's both the skills they have today and the lack of time to build the skills because they have to deal with „Old Work“ every day (Interview 6).

Even in the long term, the interviewees do not see the HR business partner carrying out the analyses with HR analytics. One interviewee cynically noted that a " new species of HR employees must first be born " (Interview 7). Even with advanced competencies in the necessary statistical methods, decentralized performance of analyses by different HR business partners is seen critically. By using different data sets and different methods to conduct the analysis, different people can come to different results: " there were three different people who gave three different results " (Interview 8). The interviewees therefore strongly emphasized the need for a central implementation of HR analytics. This can be described as a vicious circle: administrative tasks continue to dominate the task profile of HR employees and thus there is no time for further training in strategically oriented methods such as HR analytics, which should actually provide relief for operational work. Even in the long term, the interviewees do not see that HR employees will be able to conduct analyses on their own.

4.5.2 Future role model

Our analysis revealed that the HR business partner will have to take on a different role in the future than that of the analyst. The interviewees emphasized that the HR business partner must develop a stronger sense of working with analytics results in the future. Above all, the interviewees highlighted the function as a mediator and consultant between the HR department, the specialist departments and the HR analytics function. In particular, our findings showed that the implementation of HR analytics is still seen as a centralized independent function. Thereby, the task profile will change due to the closer cooperation with the HR analytics function in the sense that there will be a closer exchange between HR management and the specialist departments. In the future, HR business partners will increasingly take on an advisory role based on the analyses carried out by HR analytics. On the one hand, they should record the requirements of the specialist departments with the help of the necessary HR expertise and communicate these to the HR analytics function in a comprehensible way.

They take the requirements from the business department and then translate them into IT. And in my eyes, something like this is also missing in HR, where a business analyst takes the requirements from HR, so to speak, and then makes them available to the data scientist. That role between the different stakeholders that are involved in the process of a data-based analysis (Interview 15). We have said that we do not want this analysis as a service, but we would like to participate in the process because we also want to build up the know-how. Yes, and we work together within this framework. We personally have an interpreter function. That is, there are the data scientists who bring the methodology, who build the tool. On the other hand, there is the specialist department, which would like to know what kind of statements are contained in these free-text comments, and we as People Analytics are the interpreters between both worlds and of course also use this for us to build up the know-how (Interview 3).

Our interviewees highlighted, that the future role model must also change in such a way that the old understanding of HR work changes. Working with data and people must not be mutually exclusive in the minds of HR business partners, but must be thought of as a unit. At the present time, this is not yet possible:

Even if it's not explicitly stated: this caveat, when we, when we talk about people, we shouldn't do it in a quantitative way. In terms of feeling, that's something that resonates very often (Interview 1).

4.6 Social system: organization

4.6.1 management support.

The interviewees outlined that the role of management is a key aspect of the implementation of HR analytics. When introducing HR analytics on the part of HR, it must be ensured that management is also committed to it.

But it is usually not enough if somehow only one, yes, a sub-department head somewhere says I would like to do the whole thing. Then it fails with a sometime at the latest at the point when the whole thing is presented to the board or something else because they do not consider the whole thing so important yet (Interview 14).

Two key aspects could be identified. First, management plays an important role in legitimizing HR analytics. In-depth analyses are often initiated by management and the results are also fed back to management. However, this also clearly limits the implementation of further analysis projects.

But that is the reason why we have the backing, so to speak, for the projects that we then do. You can look at it the other way round and say that we only do the ones where we have the backing. There aren't many of them, it has to be said (Interview 1).

Secondly, one interviewee also pointed out that management's conviction must also be viewed critically. Trend topics such as data-based analyses in particular only deliver added value if they are understood in their entirety and are not just introduced because it is the latest trend.

But currently, it's really still a lot: this is a trend, I have to jump on it. Data Science in general and the whole artificial intelligence topic is so hyped and there are a lot of articles about it, which the management has picked up somewhere and then they have to do something about it, but just not this, this good understanding of what that actually means for such an HR department or where a department actually stands right now (Interview 8).

In addition to management support, the use of HR analytics also requires management to actively demand work with key figures. This means that management must demand more work with data from HR. This must be done alongside strengthening the understanding of working with data. Demanding analytics results from HR business partners thus additionally contributes to making working with HR analytics more natural for HR business partners. At this stage, the additional involvement is seen as another time factor. It should be noted, however, that according to our analysis the roles in the process of conducting HR analytics are not clearly defined, i.e. the management also does not have a clear contact person for the final analysis results.

4.6.2 Added value

One aspect that the interviewees considered particularly difficult to realize in the implementation of HR analytics is the recording of the potential added value respectively the representation of this in monetary key figures. This is particularly difficult because the causal effects of the analyses can only be clearly proven in the rarest of cases. The added value is usually justified, if at all, by time-saving or an increase in employee satisfaction. Another aspect why the added value is not yet captured is the novelty of the topic in the organizations. The interviewees were aware of the need to demonstrate added value, especially to management, but at this stage, they are focusing on conducting analyses. Projects are carried out, which are also approved by the management, since the skepticism is large here, as evidenced by the following statement:

So, the skepticism in this regard is huge. That has to be said very, very clearly. But that is the reason why we have the backing, so to speak, for the projects that we then do (Interview 1).

One interviewee critically noted that the justification and legitimization for the use of HR analytics is flimsy. Data-driven decision-making in HR management is what other corporate functions have had firmly anchored in their structures for many years, and it brings significant benefits there.

Above all, people analytics is a tool to show where inefficiencies are, the alternative to running people analytics is not running it and not knowing what's going on. That's just it, you wouldn't do that in any other area. And in other areas you wouldn't say, do we really need marketing analytics? Of course, you do. And consequently, this retroactive and block position: 'Well, does it really do anything?', I always find a bit flimsy (Interview 9).

To date, none of the interviewees has specifically established a structural process in the sense of tracking the effectiveness of the analysis results and the derived measures.

5 Systematic process perspective

After examining the results of the qualitative analysis, we also carried out an analysis of the findings from a systematic process perspective. Figure  3 shows the process flow of the implementation and application of HR analytics. Our findings reveal that HR analytics in its more structural anchoring and organizational function is not part of the HR department, but operates autonomously alongside the established HR management as a service provider for various internal customers, including the HR department. When considering the implementation of HR analytics, the first thing that stands out is that most firms tend not to involve HR managers directly in the application. Employees who are responsible for HR analytics are mainly not members of the HR department but belong to departments with a statistical background, such as data scientists, business psychologists or sociologists. Depending on the size of the HR analytics team and the given resources of the respective company, manpower and/or the knowledge from departments with data affinity, e.g. Data Science, are used.

Process model of implementation of HR analytics: status-quo

This service character of HR analytics is emphasized by many interviewees and justified, among other things, by the lack of competencies and understanding of numbers on the part of HR managers. The finding of the role of a service provider highlights the critical issue that, in particular in the collaboration with HR, the tasks of HR analytics are not clearly specified. The implications of this missing clarification of responsibilities are twofold. First, it will hinder the successful implementation of tools supporting HR analytics. Secondly, this leads to the added value of the analysis so far not being captured. It is important to define who is ultimately the recipient of the analysis results and who communicates them within the company. The results have shown that the HR analytics function on the one hand directly communicates results to the specific departments or management. The HR management receives these results partly only as information in the form of key figures in a dashboard.

A systematic process for recording the added value of the measures or the implementation of the measures does not exist. A discussion of the results on the part of HR management and the specialist department also often does not take place due to the lack of understanding and access. If we summarize statements regarding existing data, firms with existing HR controlling in particular benefit from historically grown preparatory work. Ultimately, HR analytics at the various levels of maturity often requires the same data as traditional HR controlling. The problems of data management are rather related to the technology factor, as it was pointed out here that especially outdated systems force one to deal more comprehensively with the constant data preparation. The aspect of data in the context of the implementation and application of HR analytics can thus generally be regarded as a factor that takes time but is often given in terms of quantity, insofar as the basic conditions are already present in a company. Angrave et al. ( 2016 ) highlight that the HRIS systems in use do not provide the necessary analysis capabilities. This point should be viewed critically, as our results have shown that in the short to medium term, the HR business partner will not be the user of the analysis either. It is also not expected in the long term that the role of the HR business partner will be sharpened in such a way that it can perform and understand data-based analyses. Rather, interviewees see the HR business partner involved in the sense that HR expertise from an HR analytics perspective is needed for targeted analyses. The technology question in this sense does not arise at all from this point of view of the methodological possibilities, but rather how the analysis results can be visually represented by another tool such as Tableau. The actual analysis, carried out by the HR analytics function, will draw on its methodological knowledge and adept programs with its competencies in the analysis.

6 Discussion

Falletta and Combs ( 2021 ) have noted that despite the interest in the topic of HR analytics, the actual knowledge about it is still in its infancy. This starts with the lack of a common definition and ends with a lack of knowledge about the processes of applying and implementing HR analytics in an intra-organizational context (Falletta and Combs 2021 ; Greasley and Thomas 2020 ). The aim of our research was to examine the implementation and application of HR analytics in firms. By applying a socio-technical approach we have developed a theoretical framework which integrates four areas impacting the implementation and application of HR analytics. This framework guided our qualitative research study, which resulted in a more nuanced understanding of the facets of HR analytics as well as its implementation process. The conclusions we have reached from the results of our study will be discussed as follows: first, we will reflect the results of our qualitative study in light of the socio-technical system perspective. From a practical perspective, we then propose a process model for the future application of HR analytics.

6.1 Theoretical implications

In sum, our results showed that there is a common understanding of the future use of HR analytics. This implies the improvement of HR-related decision-making by using a data-driven approach, which aims at achieving strategic business goals. However, a lot of HR analytics analyses are currently in the early stages and are partly project-based or prototype-based. Thereby, firms still face many challenges. This is in line with the findings of Fernandez and Gallardo-Gallardo ( 2021 ) emphasizing that firms need to overcome organizational barriers with regard to HR analytics. Our study contributes to this important endeavor by examining the barriers of the social and the technical system:

Our first observation relates to the use of available data and technologies for the application of HR analytics. The availability of data and data integration is considered one of the most important factors in the implementation of HR analytics (Halper 2014 ; Pape 2016 ). Our findings confirm this important factor. It should be noted, however, that the interviewees see data integration as a rather operational task, which primarily requires time resources rather than competencies. The provision of the necessary data depends primarily on the level of technologization and the HRIS system used specifically in HR. We did not find any issues regarding the use of external data. However, this may also be due to the fact that many of the respondents are just beginning to explore their options and have not yet conducted analyses that require a deeper data set. It can be noted that the current level of maturity does not meet the demands of some authors who call for a holistic HR analytics function that also includes departments such as finance, production, etc. (Rasmussen and Ulrich 2015 ; Marler and Boudreau 2017 ; McIver et al. 2018 ). The picture is similar for the technology used for analysis: our interviewees have advanced skills, so applying the necessary skills is not a problem. However, it turned out that the issue of technology is much more related to the delivery and visualization of the analytics results rather than to the actual analysis tools. A similar conclusion was also reached by van den Hauvel and Bondarouk ( 2017 ). The authors emphasize that HR analytics goes beyond mere analysis and requires convincing visualization and presentation. (van den Hauvel and Bondarouk 2017 ). Our findings have shown that there is more of a concern with creating awareness that the HR business partner is working with these analytics results and using the dashboards provided (Vargas et al. 2018 ). It might be argued that the previous data and technology challenges were viewed from the perspective of HR business partners as users of HR analytics. However, our study has shown that they are not the actual people who perform advanced analytics using HR analytics. The HR business partner is much more understood as a customer in this process (Jörden et al. 2021 ).

We now turn to the discussion of the results of the social system, i.e. the analysis of the role of the HR business partner and intra-organizational context. Previous research has identified two different approaches to embedding HR analytics in the organization. On the one hand, HR analytics is seen as an independent function that uses the potential of data scientists (Rasmussen and Ulrich 2015 ). On the other hand, it is argued that an integration of HR analytics in HR management is worthwhile (Bassi 2011 ; Falletta and Combs 2021 ).

Our results have shown that the preferred path of the firms in our study is the former and that a new stand-alone HR analytics function is emerging alongside the HR function itself. This may be due to the fact that analytical skills in HR are not sufficient to apply HR analytics independently (Angrave et al. 2016 ; Marler and Boudreau 2017 ). Although our results have shown that HR itself can also be identified as the initiator of HR analytics, the function does not emerge within HR but is carried out by individuals who inherently bring the necessary competencies. This approach can be promising but runs the risk of leaving out HR managers who have been active to date. Ultimately, this can lead to competencies being built up in silos in the long term and not being taught throughout the HR sector. To prevent this, a holistic approach is needed. HR analytics should not be considered as a stand-alone function of the HR value chain but should map it across the entire functions of the HR value chain.

Another important factor that plays a role is the recording of the added value of HR analytics. Our findings have shown that many analytics results are communicated to the respective addressees, e.g. the specific departments, HR or management without an accurate evaluation of the effectiveness or efficiency of these measures afterward. However, one of the core objectives of HR analytics is to improve organizational productivity and employee experience (Tursunbayeva et al. 2018 ). This lack of evaluation also means that HR management's own strategic strengths fall short of expectations and thus the reasons for the legitimacy of HR analytics cannot be optimally communicated to management (Bassi et al. 2012 ). As shown in our status-quo process model (see Fig.  3 ), HR analytics is currently acting autonomously as a service provider. This finding is in line with the observations of Jörden et al. ( 2021 ), who also found in an ethnological study of a people analytics team that HR analytics was “primarily driven and restricted by customer requirements, and as a consequence PA as a specialist professional HR practice was undermined by a lack of managerial commitment to technical quality “(Jörden et al. 2021 : 11). Especially management as a customer is a double-edged sword: our study has shown that the legitimization of HR analytics is a decisive factor in the implementation of HR analytics. On the other hand, Jörden et al. ( 2021 ) see management as a critical factor in this respect, as it can also undermine the possibilities of HR analytics by only carrying out analyses that are demanded by management.

Moreover, we know little about which measures are ultimately implemented by the individual addressees from the derived analysis results (Ellmer and Reichel 2021 ). The lack of evaluation of the derived measures leads to the fact that the seemingly relevant business context of the analyzed issue cannot be clearly evidenced (McIver et al. 2018 ). In the future, HR analytics must have more legitimacy grounds than those of management. This means closer cooperation between firmly integrated functional areas and HR so that HR analytics does not become an end in itself and can bring the promised added value (Rasmussen and Ulrich 2015 ).

In summary, it can be stated that there is a need to better understand HR analytics from a process perspective. This implies to define the different internal stakeholders which are involved in the HR analytics process and to cover their respective ideas and wishes. In line with Ellmer and Reichel ( 2021 ) our findings show that the role of the HR business partner needs to be defined more clearly in the future. Also, the required understanding and affinity for the work with data and data-based analyses can only succeed if the scope of responsibility of the HR department is clearly defined. A sharpening of the role of HR can help to clarify whether HR analytics promises the actual added value, i.e. an increase in HR’s strategic alignment in the organizational context (Greasley and Thomas 2020 ). Still, it is difficult to implement a purely autonomous execution of HR analytics without the involvement of traditional human resource management (Bassi 2011 ). Thus, the targeted communication of the necessary competencies and the definition of the areas of responsibility for HR analytics is a necessary step that firms will have to take to ensure the effective application of HR analytics. In this regard, Falletta and Combs ( 2021 ) don’t position HR analytics as a separate function but argue that it should be located directly in HR management. This contradicts previous research which concluded that HR analytics should be a permanent function (Rasmussen and Ulrich 2015 ; Ulrich and Dulebohn 2015 ).

6.2 Managerial implications

In light of our findings, we argue that in order to strengthen the practical implementation and application of HR analytics the following aspects can be defined as adjusting screws from a process perspective: At a social level, there is a need for a clearer clarification of roles in the intra-organizational process of HR analytics. At the technical level, one needs to be aware that adapted software solutions have to fit the respective competencies of the HR business partners. Based on this we propose a process model for the future application of HR analytics (see Fig.  4 ). Our study has shown that it is not necessary for an HR business partner to have the skills to carry out analyses independently. Rather it is crucial to develop an awareness and speak the language required to understand these analyses and discuss them with the other departments in the next step. A discussion of whether the users of HR analytics have a statistical and business background or rather a social, behavioral and organizational sciences background (Falletta and Combs 2021 ) is not expedient. The following applies here: many solutions lead to the goal; ultimately, it is important that the analyses are based on valid methods and are goal-oriented from the perspective of HR management. In order for an HR business partner to be able to work with these analysis results, more user-friendly software solutions will be needed in the future. Again, it is not a question of carrying out the analysis itself, but rather of providing the necessary information in a targeted manner in order to enter into an exchange with the specialist departments. The role of the HR business partner as a future consultant is seen as particularly important as the previous way of communicating the results of the analyses is not precisely defined: If it is possible for the HR department to make decisions in cooperation with the specialist department on the basis of analysis results, this is also reflected in the communication with management. This in turn has a positive effect on management's attitude toward HR. In the long run, the closer integration of HR analytics and the HR business partners can lead to achieving the actual goal of strengthening the strategic role of HR. For these reasons, we advocate a multidimensional role model for the application of HR analytics.

Supposed process model for the future of implementation of HR analytics

7 Limitations

The study has some limitations. First, the results are based on the conduct of qualitative research. This methodological approach of qualitative research naturally entails some limitations. The results obtained cannot be generalized to a broader population with the same degree of generalizability. However, statistical analysis of the results is also not intended, as the goal of qualitative research should be to gain new insights based on experience and detailed descriptions. We recommend that our proposed model of HR analytics implementation be tested in the future using descriptive and observational studies. Since HR analytics is an interface function between HR and business informatics, one possible approach would be to apply design science research from business informatics. The approach starts from an application-oriented problem, on the basis of which an IT artifact is created and tested in several iterative steps. Future research could target approaches here and investigate different IT artifacts to explore the implementation and application of HR analytics more deeply empirically.

Second, the findings of our study are based on interviews with HR analytics experts. These experts have been chosen according to a clearly defined set of criteria. Given that the implementation and application of HR analytics is still in its infancy in most of the firms, the knowledge base of interviewees is at a beginner or mediocre level. To gain more insights, we recommend to analyze the ongoing development of HR analytics and its structural positioning within organizations once the application and implementation have become more deeply established in firms. This would enable further exploration of the role of the HR business partner as well as identify strategies of how to close the gap between HR analytics and HR function. In addition, we did not examine any other stakeholders involved in the process (e.g., management, HR business partners or specialist departments) as part of this study. In order to get a more holistic picture of the model proposed by us, we recommend conducting further interviews with these groups as well. As a final note, our study was only conducted in German-speaking countries, so the use of HR analytics must be reflected under the applicable data protection aspects.

Third, as mentioned at the beginning, the COVID 19 pandemic has given a new boost to the topic of HR analytics. The sharp increase in the use of digital technologies and the changes in working conditions offer a wide range of new opportunities for analyzing HR issues. Since this study took place precisely during the pandemic and the impact of the pandemic was not a focus, future research should investigate the extent to which the new circumstances influence the implementation and application of HR analytics. In addition, it should be quantitatively investigated to what extent HR analytics could be refined based on the multitude of new types of data sources resulting from the new types of working conditions.

8 Conclusion

Firms have recognized the opportunities presented by HR analytics; however, it is challenging for HR to convert their available data (sources) into meaningful strategical value. Moreover, scant research has explored how the implementation and application of HR analytics is achieved. This study provides one of the first attempts to examine the socio-technical aspects that underline the process of HR analytics. The results of this study contribute to the existing literature by showing that the function of HR analytics needs to be reconsidered. Also, it encourages future research to dive deeper into the variety of contextual and process conditions strengthening or weakening the value of HR analytics.

Acito F, Khatri V (2014) Business analytics: why now and what next? Bus Horiz 57(5):565–570. https://doi.org/10.1016/j.bushor.2014.06.001

Article   Google Scholar  

Andersen MK (2017) Human capital analytics: the winding road. J Organ Eff 4(2):133–136. https://doi.org/10.1108/JOEPP-03-2017-0024

Angrave D, Charlwood A, Kirkpatrick I, Lawrence M, Stuart M (2016) HR and analytics: why HR is set to fail the big data challenge. Hum Resour Manag J 26(1):1–11. https://doi.org/10.1111/1748-8583.12090

Aral S, Brynjolfsson E, Wu L (2012) Three-way complementarities: performance pay, human resource analytics, and information technology. Manage Sci 58(5):913–931. https://doi.org/10.1287/mnsc.1110.1460

Bassi L (2011) Raging debates in HR analytics. People Strategy 34(2):14–18

Google Scholar  

Bassi L, Carpenter R, McMurrer D (2012) HR analytics handbook. Reed Business, Amsterdam

Belizón MJ, Kieran S (2021) Human resources analytics: a legitimacy process. Hum Resour Manag J. https://doi.org/10.1111/1748-8583.12417

Biazzo S (2002) Process mapping techniques and organisational analysis: lessons from sociotechnical system theory. Bus Process Manag J 8(1):42–52. https://doi.org/10.1108/14637150210418629

Boudreau J, Cascio W (2017) Human capital analytics: why are we not there? J Organ Eff 4(2):119–126. https://doi.org/10.1108/JOEPP-03-2016-0029

Bryce V, McBride NK, Cunden M (2022) Post-COVID-19 ethics of people analytics. J Inf Commun Ethics Soc. https://doi.org/10.1108/JICES-09-2021-0096

Cascio WF, Montealegre R (2016) How technology is changing work and organizations. Annu Rev Organ Psychol 3:349–375. https://doi.org/10.1146/annurev-orgpsych-041015-062352

Cherns A (1976) The principles of sociotechnical design. Hum Relat 29(8):783–792. https://doi.org/10.1177/001872677602900806

Coakes E (2002) Socio-technical thinking-an holistic viewpoint. In: Coakes E, Willis D, Clarke S (eds) Sociotechnical and human cognition elements of information systems. Springer, London, pp 1–4

Coco CT, Jamison F, Black H (2011) Connecting people investments and business outcomes at Lowe’s: using value linkage analytics to link employee engagement to business performance. People Strategy 34(2):28–33

Côrte-Real N, Oliveira T, Ruivo P (2017) Assessing business value of big data analytics in European firms. J Bus Res 70:379–390. https://doi.org/10.1016/j.jbusres.2016.08.011

Douthitt S, Mondore S (2014) Creating a business-focused HR function with analytics and integrated talent management. People Strategy 36(4):16–21

Earley CE (2015) Data analytics in auditing: opportunities and challenges. Bus Horiz 58(5):493–500. https://doi.org/10.1016/j.bushor.2015.05.002

Ellmer M, Reichel A (2021) Staying close to business: the role of epistemic alignment in rendering HR analytics outputs relevant to decision-makers. Int J Hum Resour Man 32(12):2622–2642. https://doi.org/10.1080/09585192.2021.1886148

Falletta SV (2014) In search of HR intelligence: evidence-based HR Analytics practices in high performing companies. People Strategy 36(4):28–37

Falletta SV, Combs WL (2021) The HR analytics cycle: a seven-step process for building evidence-based and ethical HR analytics capabilities. J Work-Appl Manag 13(1):51–68. https://doi.org/10.1108/JWAM-03-2020-0020

Fernandez V, Gallardo-Gallardo E (2021) Tackling the HR digitalization challenge: key factors and barriers to HR analytics adoption. Compet Rev 31(1):162–187. https://doi.org/10.1108/CR-12-2019-0163

George G, Haas MR, Pentland A (2014) Big data and management. Acad Manage J 57(2):321–326. https://doi.org/10.5465/amj.2014.4002

Ghasemaghaei M (2018) Improving organizational performance through the use of big data. J Comput Inf Syst. https://doi.org/10.1080/08874417.2018.1496805

Giuffrida M (2014) Unleashing the power of talent analytics in federal government. Public Manager 43(3):7–10

Graebner ME, Martin JA, Roundy PT (2012) Qualitative data: cooking without a recipe. Strateg Organ 10(3):276–284. https://doi.org/10.1177/1476127012452821

Greasley K, Thomas P (2020) HR analytics: the onto-epistemology and politics of metricised HRM. Hum Resour Manag J 30(4):494–507. https://doi.org/10.1111/1748-8583.12283

Halper F (2014) Predictive analytics for business advantage. https://vods.dm.ux.sap.com/previewhub/ITAnalyticsContentHubANZ/downloadasset.2014-03-mar-17-21.predictive-analytics-for-business-advantage-pdf.pdf Accessed from 15 Oct 2021

Huselid MA (2018) The science and practice of workforce analytics: introduction to the HRM special issue. Hum Resour Manag 57(3):679–684. https://doi.org/10.1002/hrm.21916

Hyde KF (2000) Recognising deductive processes in qualitative research. Qual Mark Res 3:82–90. https://doi.org/10.1108/13522750010322089

Jaffee D (2001) Organization theory: tension and change. McGraw-Hill, New York

Jeske D, Calvard T (2020) Big data: lessons for employers and employees. Empl Relat 42(1):248–261. https://doi.org/10.1108/ER-06-2018-0159

Jörden NM, Sage D, Trusson C (2021) It’s so fake’: identity performances and cynicism within a people analytics team. Hum Resour Manag J. https://doi.org/10.1111/1748-8583.12412

King KG (2016) Data analytics in human resources: a case study and critical review. Hum Resour Dev Rev 15(4):487–495. https://doi.org/10.1177/1534484316675818

Kniffin KM, Narayanan J, Anseel F, Antonakis J, Ashford S, Bakker AB (2020) COVID-19 and the workplace: Implications, issues, and insights for future research and action. https://doi.org/10.1037/amp0000716

Kryscynski D, Reeves C, Stice-Lusvardi R, Ulrich M, Russell G (2018) Analytical abilities and the performance of HR professionals. Hum Resour Manage 57(3):715–738. https://doi.org/10.1002/hrm.21854

Lawler E, Boudreau JW (2015) Global trends in human resources management: A twenty- year analysis. Stanford University Press, Stanford

Levenson A (2011) Using targeted analytics to improve talent decisions. People Strategy 34:34–43

Levenson A, Fink A (2017) Human capital analytics: too much data and analysis, not enough models and business insights. J Organ Eff 4(2):145–156. https://doi.org/10.1108/JOEPP-03-2017-0029

Marler JH, Boudreau JW (2017) An evidence-based review of HR analytics. Int J Hum Resour Man 28(1):3–26. https://doi.org/10.1080/09585192.2016.1244699

Maucher I, Paul H, Rudlof C (2002) Modellierung in soziotechnischen systemen. In: Desel J, Weske M (eds) Prozessorientierte methoden und werkzeuge für die entwicklung von informationssystemen. Gesellschaft für Informatik, Bonn, pp 128–137

Mayring P (2014) Qualitative content analysis: theoretical foundation. Basic procedures and software solution. Klagenfurt, Beltz

McCartney S, Fu N (2022) Promise versus reality: a systematic review of the ongoing debates in people analytics. J Organ Eff. https://doi.org/10.1108/JOEPP-01-2021-0013

McIver D, Lengnick-Hall ML, Lengnick- Hall CA (2018) A strategic approach to workforce analytics: integrating science and agility. Bus Horiz 61(3):397–407. https://doi.org/10.1016/j.bushor.2018.01.005

Minbaeva D (2018) Building credible human capital analytics for organizational competitive advantage. Hum Resour Manage 57(3):701–713. https://doi.org/10.1002/hrm.21848

Mondare S, Douthitt S, Carson M (2011) Maximizing the impact and effectiveness of HR Analytics to drive business outcomes. People Strategy 34:20–27

Mumford MD (2003) Where have we been, where are we going? Taking stock in creativity research. Creat Res J 15(2–3):107–120. https://doi.org/10.1080/10400419.2003.9651403

Pape T (2016) Prioritising data items for business analytics: framework and application to human resources. Eur J Oper 252(2):687–698. https://doi.org/10.1016/j.ejor.2016.01.052

Peeters T, Paauwe J, Van De Voord K (2020) People analytics effectiveness: developing a framework. J Organ Eff 7(2):203–219. https://doi.org/10.1108/JOEPP-04-2020-0071

Rasmussen T, Ulrich D (2015) Learning from practice: how HR-Analytics avoids being a management fad. Organ Dyn 44(3):236–242. https://doi.org/10.1016/j.orgdyn.2015.05.008

Shrivastava S, Shaw JB (2003) Liberating HR through technology. Hum Resour Manage 42(3):201–222. https://doi.org/10.1002/hrm.10081

Snell SA, Stueber D, Lepak DP (2002) Virtual HR departments: getting out of the middle. In: Heneman RL, Greenberger DB (ed) Human resource management in virtual organizations. pp 81–101

Trist EL, Bamforth KW (1951) Some social and psychological consequences of the Longwall method of coal-getting: an examination of the psychological situation and defences of a work group in relation to the social structure and technological content of the work system. Hum Relat 4(1):3–38. https://doi.org/10.1177/001872675100400101

Tursunbayeva A, Di Lauro S, Pagliari C (2018) People analytics—A scoping review of conceptual boundaries and value propositions. Int J Inform Manage 43:224–247. https://doi.org/10.1016/j.ijinfomgt.2018.08.002

Ulrich D (1997) HR of the future: conclusions and observations. Hum Resour Manage 36:175–179

Ulrich D, Dulebohn JH (2015) Are we there yet? What’s next for HR? Hum Resour Manage R 25(2):188–204. https://doi.org/10.1016/j.hrmr.2015.01.004

van den Heuvel S, Bondarouk T (2017) The rise (and fall?) of HR-analytics. J Organ Eff 4(2):157–178. https://doi.org/10.1108/JOEPP-03-2017-0022

van der Togt J, Rasmussen TH (2017) Toward evidence-based HR. J Organ Eff 4(2):127–132. https://doi.org/10.1108/JOEPP-02-2017-0013

Vargas R, Yurova YV, Ruppel CP, Tworoger LC, Greenwood R (2018) Individual adoption of HR analytics: a fine grained view of the early stages leading to adoption. Int Hum Resour Man 29(22):3046–3067. https://doi.org/10.1080/09585192.2018.1446181

Wirges F, Neyer, AK, Kunisch M (2020) HR-Studie 2020: So steht es um die Digitalisierung der Personalarbeit: Inwiefern Human Resources 4.0 bereits Realität ist und welche Potenziale noch ungenutzt sind. Studienreihe der forcont business technology gmbh

Download references

Open Access funding enabled and organized by Projekt DEAL.

Author information

Authors and affiliations.

Chair of Human Resources Management & Business Governance, Martin-Luther-University Halle-Wittenberg, Halle, Germany

Felix Wirges & Anne-Katrin Neyer

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Felix Wirges .

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

1.1 Interview guide

1.1.1 application.

What do you understand by the term HR analytics?

Where does it differ from classic HR controlling?

What concrete added value does the use of HR analytics offer?

Can you quantify the added value using key figures? (e.g. resource savings, quality of recruitment, more transparency, etc.).

If not: How would you describe the added value subjectively?

What has changed in the HR department as a result of using HR analytics? What has changed from an overall organizational perspective?

Can you identify specific areas where data-driven decisions have increased?

What was the cause of this exact area being supported by data analytics? What exactly has changed about the process? How did this decision take place in the past? Have there been any advantages or disadvantages to this?

What problems were/are there in general with the evaluation of HR data?

On the technical/data level

On the evaluation level

What are the data sources for the HR data used?

How is this data systematically collected?

Who has access to this data?

Are there any problems regarding interfaces in data collection?

Which specific questions are answered with the help of the data analysis? Why these in particular?

Which data are necessary for this?

How do you collect and analyze specifically qualitative data?

Do you include external data in the analyses, if so which ones? What difficulties do you encounter in doing so?

1.3 Technology

What software do you currently use to analyze HR data? Dedicated HR analytics software or individual tools in different application areas?

What skills does it take to use this software/tool? Did you already have these skills or were they obtained elsewhere?

What difficulties are encountered with this software/tool? How could these be improved?

1.4 HR business partner

Who has been the primary initiator of HR data analysis to date? Why?

What does the power user of HR data analysis look like? Individual person or a data team?

What are the advantages and disadvantages of this?

What other stakeholders are involved in the process?

At which interfaces do difficulties arise? Which ones?

Has the use of HR analytics changed the collaboration with other departments/areas? In what form? If not, would it have to change from your point of view?

1.5 Organization

What organizational difficulties have arisen in the analysis of HR data? How were/are they dealt with?

What difficulties have arisen in the analysis of HR data on the part of the employees? How was/is this dealt with?

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Wirges, F., Neyer, AK. Towards a process-oriented understanding of HR analytics: implementation and application. Rev Manag Sci 17 , 2077–2108 (2023). https://doi.org/10.1007/s11846-022-00574-0

Download citation

Received : 31 March 2021

Accepted : 10 July 2022

Published : 18 August 2022

Issue Date : August 2023

DOI : https://doi.org/10.1007/s11846-022-00574-0

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• HR analytics
• People analytics

JEL Classification

Advertisement

• Find a journal
• Publish with us
• Track your research

HR analytics-as-practice: a systematic literature review

Journal of Organizational Effectiveness: People and Performance

ISSN : 2051-6614

Article publication date: 21 December 2023

Human resource analytics (HRA) is an HR activity that companies and academics increasingly pay attention to. Existing literature conceptualises HRA mostly from an objectivist perspective, which limits understanding of actual HRA activities in the complex organisational environment. This paper therefore draws on the practice-based approach, using a novel framework to conceptualise HRA-as-practice.

Design/methodology/approach

The authors conducted a systematic literature review of 100 academic and practitioner-oriented publications to analyse existing HRA literature in relation to practice theory, using the “HRA-as-practice” frame.

The authors identify the main practices involved in HRA, by whom and how these practices are enacted, and reveal three topics in nomological network of HRA-as-practice: HRA technology, HRA outcomes and HRA hindrances and facilitators, which the authors suggest might actualize enactment of HRA practices.

Practical implications

The authors offer HR function and HR professionals a basic ground to evaluate HRA as a highly contextual activity that can potentially generate business value and increase HR impact when seen as a complex interaction between HRA practices, HRA practitioners and HRA praxis. The findings also help HR practitioners understand multiple factors that influence the practice of HRA.

Originality/value

This systematic review differs from the previous reviews in two ways. First, it analyses both academic and practitioner-oriented publications. Second, it provides a novel theoretical contribution by conceptualising HRA-as-practice and comprehensively compiling scattered topics and themes related to HRA.

• Human resource analytics
• HR analytics
• Practice theory
• HR practices
• HR practitioners

Espegren, Y. and Hugosson, M. (2023), "HR analytics-as-practice: a systematic literature review", Journal of Organizational Effectiveness: People and Performance , Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/JOEPP-11-2022-0345

Emerald Publishing Limited

Copyright © 2023, Yanina Espegren and Mårten Hugosson

Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode

1. Introduction

Human resource analytics (HRA) is a human resource (HR) activity that has recently attracted growing interest among companies and public organisations. HRA has broadly been seen as the collection, analysis and reporting of data to inform people-related decisions and improve individual and organisational outcomes ( Fernandez and Gallardo-Gallardo, 2021 ).

The interest in HRA has evolved mainly through HR management (HRM) experts and business consultants, who portray HRA as creating numerous advantages for organisations and the HR profession. These advantages include, e.g. improved rigour of HR decisions, increased credibility and strategic value of the HR function, enhanced competitive advantage and business success that can be achieved through a better understanding of the organisational workforce ( Davenport et al. , 2010 ; Huselid, 2018 ; Deloitte, 2021 ).

However, despite the enhanced popularity of the topic, the academic field of HRA is still reported as nascent, the research, though growing, is rather scarce, lacking a unifying understanding of HRA practice, possibly due to the continuously increasing development and evolution of the HRA scope ( van den Heuvel and Bondarouk, 2017 ; Margherita, 2022 ). HRA literature remains scattered, with too few consistent frameworks, and its current state can be described as “wild, wild west” ( Levenson and Fink, 2017 ).

In an attempt to grasp the scope of HRA and reflect on its early development, several literature reviews have been conducted in the field ( Marler and Boudreau, 2017 ; Tursunbayeva et al. , 2018 , 2022 ; Ben-Gal, 2019 ; Fernandez and Gallardo-Gallardo, 2021 ; Margherita, 2022 ; Qamar and Samad, 2022 ; McCartney and Fu, 2022 ; Jiang and Akdere, 2022 ; Giermindl et al. , 2022 ). Despite the difference in purposes and research questions, these literature reviews are similar in their understanding of HRA from an objectivist perspective, namely, as something – process or tool – that organisations have or lack. This view is, however, useful but somewhat deficient and fragmented in its depiction of the complexity of organisational reality and focuses more on what should be done to succeed with HRA rather than on the actual activities within organisations. Thus, as also indicated by Jiang and Akdere (2022) , there is a clear need for theoretical development in the field as it stands at the present.

The purpose of this paper is therefore to draw on the alternative perspective originated in the practice-based ontology and reconceptualise HRA using a new framework of HRM-as-practice. This conceptualisation has three dimensions: HRA practices, HRA praxis and HRA practitioners.

HRM-as-practice framework was proposed by Björkman et al. (2014) as a holistic and more dynamic approach to understand and innovatively examine the general practice of HRM. The framework addresses intersections between the three components of general HRM practice – practices, praxis and practitioners – by asking, “ What” general practices does HRM involve? “ Who” are the HRM practitioners? “H ow” do HRM practitioners enact HRM practices? Applying this approach, we intend to answer the following research questions: (1) What practices constitute HRA? (2) How are these practices enacted in organisations? (3) What actors are involved in the enactment of the practices? And, finally, (4) What connects HRA practices, their enactment and HRA practitioners into a coherent model of HRA-as-practice?

To answer the research questions, we conducted a literature assessment of academic and practitioner-oriented articles. Considering the general lack of empirical academic research in the field ( Edwards et al. , 2022 ), this methodological approach allows for a wider coverage of material for the analysis, particularly with our special focus on the practice of HRA.

This study thus provides an overview of the HRA field from a practice perspective and aims to contribute to the theoretical understanding of HRA by constructing it as a coherent and holistic concept. Moreover, it also compiles a nomological network of HRA-as-practice, adding more recent and ample findings to the topics and concepts that exist around the practice of HRA and influence and are influenced by it. Practically, it offers HR functions and HR professionals an overview of how HRA operates as a practice within organisations. It also provides a basis for evaluating the conditions enabling and moderating HRA as a possible solution to generate business value and increase HR impact due to a better understanding of the important features of HRA and what is implied for its useful enactment.

The paper is structured as follows: first, the practice theory is briefly introduced, followed by a description of the methodology. Next, the results of the literature review are presented and discussed.

2. Theoretical approach: HRA-as-practice

For our analysis, we have chosen to elaborate on a framework proposed by Björkman et al. (2014) and their concept of HRM-as-practice, which in turn leans on the strategy-as-practice discourse ( Whittington, 2006 ; Jarzabkowski et al. , 2007 ). By arguing that organisational phenomena do not exist until they are enacted in practice, practice theory particularly aims to overcome the structure-agency dualism, implying that both individual doings and structural influences only acquire meaning when they manifest themselves in practice ( Nicolini, 2012 ). These ideas emerged from seminal works in sociology (e.g. Giddens, 1984 ) underpinning all practice theories and have lately been used for studying different areas of business administration and management, such as organisational learning and knowing ( Tsoukas, 1996 ), strategy ( Whittington, 1996 ), technology ( Orlikowski, 2000 ), accounting ( Ahrens and Chapman, 2006 ) and HRM ( Björkman et al. , 2014 ).

Central to the ideas of strategy-as-practice are three interrelated concepts suggesting that studying social practice should focus on practices, praxis , practitioners and intersections between them ( Whittington, 2006 ; Jarzabkowski et al. , 2007 ). Integrating these three concepts allows for a coherent depiction and understanding of a social phenomenon and how it evolves. The conceptualisation is ontologically rooted in the duality identified by Giddens (1984) , when structures and actors simultaneously influence and are influenced by each other.

The HRM-as-practice draws on these ideas and adapts the concept to the HRM field. The conceptualisation involves the three basic categories, contextually defining them as HRM practices, HRM praxis and HRM practitioners ( Björkman et al. , 2014 ). The “three Ps” are thus in line with the general practice theories, seen as inseparable, interconnected and difficult to distinguish since practices are “pertinent” when enacted by practitioners in a certain context ( Whittington, 2006 ). The model calls for an extended assessment of what links these elements in their intersections, constituting the practice of HRM as it is enacted. According to the underlying theory, particular practices are entangled with other practices, creating the problem of making clear distinctions between them. In other words, the practice theory tradition emphasises that practices are entangled in bundles, embodied in practitioners and enacted by them ( Jarzabkowski et al ., 2016 ). Thus, the distinction used in this article is basically an exercise of abstraction from a wider context of practices. It is an analytical choice intended as a necessary simplification for a better understanding of the practice of HRA.

The definition of practices varies in practice literature depending on the field of application. Björkman et al. (2014) understand practices as tools, norms, processes and procedures and traditionally exemplify HRM practices as HRM routines and techniques that ensure implementation of HRM policies, e.g. high-performance HRM practices. Since there is no traditional agreement on HRA practices due to the current evolution and development of such activities, we choose to depict them in line with the more broad understanding in the strategy-as-practice tradition as something that is “done” ( Whittington, 2006 ), the supposed or routinised activities that can be “diverse and variable” and combined and adjusted depending on their utilisation in a particular context ( Jarzabkowski et al. , 2007 ). In other words, HRA practices are seen as recognised abstract general activities that principally construct HRA.

Enacted HRA practices are situated activities, which we call HRA praxis . Jarzabkowski et al. (2007) discuss praxis as the flow of actual activities that are situated, socially accomplished and consequential. In line with this definition, we understand HRA praxis as an actual activity, representing how people “go about things” when they perform HRA practices. If HRA practices are more generally recognised patterns and principles about what is included in HRA work, HRA praxis is the actual work, representing how the abstract activities are enacted in real situations.

HRA practitioners are the actors who enact, construct and reconstruct HRA practices through their actual activities. The practitioners are seen as the “prime movers” who perform the actual work ( Whittington, 2006 ). In this study, we understand HRA practitioners as those who are directly involved in the enactment of HRA practices. The strategy-as-practice approach, although recognising both internal and external actors who impact strategy-making, often tends to focus on the individual actors and their agency. Instead, Björkman et al. (2014) follow HRM tradition, distinguishing between individual and collective actors. In this exploratory study, we adhere to a broader perspective and are interested in revealing all possible practitioners involved in HRA, both individual and collective, to get a holistic picture.

Finally, we are also interested in how the above-mentioned “three Ps”, HRA practices, HRA praxis and HRA practitioners, are actualised and connected into one whole practice of HRA. According to Björkman et al. ’s (2014) model, the three Ps are de facto entangled, and their interconnections are of prime importance for the model because HRM-as-practice manifests itself in these intersections. Björkman et al. (2014) , although stressing importance of the interconnections, neither provide any detailed descriptions of what exactly happened there nor introduce any defined entities that influence and are influenced by the practice of HRA and its elements. Instead, Björkman et al. (2014) suggest several potentially interesting research questions that arise from the intersections without being rigid guidelines and might be modified by future HRM research. Based on the elusive and, to a certain degree, obscure nature of the interconnections, we are interested in revealing what topics and concepts in the nomological network of HRA-as-practice might actualise intersections between HRA practices, HRA praxis and HRA practitioners and integrate them into a coherent whole.

3. Methodology

A systematic literature review has been conducted to analyse relevant HRA literature. The literature review process used here followed the four steps: (1) developing a review protocol; (2) searching for the literature; (3) selecting the studies for review; and (4) summarising the evidence ( Boell and Cecez-Kecmanovic, 2015 ). A PRISMA flow diagram was used to document the search and selection of the studies for the review ( Moher et al. , 2009 ). A summary of this process is depicted in Figure 1 .

According to the review protocol, three databases were searched to identify articles for the review: Scopus, Web of Science and Business Source Complete. Scopus and Web of Science were chosen as the two most well-known interdisciplinary research databases with a wide coverage of academic articles ( Chadegani et al. , 2013 ). Scopus and Web of Science each provide access to more than 80 million records in more than 21,000 journals. The database Business Source Complete, as one of the leading databases in the field of business and management studies and with access to more than 4,000 high-profile journals, was chosen as a complementary source.

Although HRA is found to be the most frequently used term for the studied phenomena ( Margherita, 2022 ), it is still not recognised as an exclusive search term in all academic disciplines ( Edwards et al. , 2022 ). The existing literature reports numerous synonyms of HRA ( Marler and Boudreau, 2017 ; Fernandez and Gallardo-Gallardo, 2021 ). In line with this condition, the following terms were applied when searching for source titles, abstracts and keywords: HR analytics, human resource analytics, talent analytics, workforce analytics, people analytics, human capital analytics and employee analytics. The usage of the different names is explained by the emergent nature of the HRA phenomenon ( Margherita, 2022 ). This paper disregards the potential semantic differences among the different labels and uses HRA as a common term for all the synonyms. The three databases were searched in November 2021 for publications published between 2010 and 2021. The starting year was chosen based on the previously reported observation that the number of HRA articles noticeably increased after 2010 and were almost non-existent before that date ( Marler and Boudreau, 2017 ).

The search was limited by type of publication, language and subject area. We included articles from journals written in English in the field of business and management. This led to the identification of 301 publications. After removing duplicates, 202 records remained. A screening process of abstracts at this point resulted in the exclusion of a further 83 records for the following reasons: the content was not relevant for the HRA topic (45); the content was of questionable quality because it was published in a journal or by a publisher listed in Beall’s list of potentially predatory journals and publishers (21); the article was an editorial (7), a book review (3), a summary of other studies (3), an internal university publication (2), an executive interview (1), or a review of conferences (1). After identification and selection, all the 119 full-text articles were read. An additional 11 were excluded based on their content because the focus was on other topics and HRA was only marginally mentioned in the text. Eight literature review articles were also excluded from the analyses because of the prime interest in empirical findings in connection to HRA. This resulted in a final total of 100 articles, which formed the basis for our analysis. A full list of articles is included in Appendix .

Among the 100 articles selected for analysis, it is possible to clearly distinguish between two broad groups. The first one comprises 53 traditional academic articles. The other group comprises 47 articles, mostly aimed at the practitioner audience. They are shorter than academic publications, do not necessarily deploy scientific methods, often have a viewpoint character or are case studies and are often published by practitioners in trade journals. We labelled these articles “practitioner-oriented”. The reason we have included practitioner-oriented publications is because they are directly related to the topic of HRA-as-practice and cover relevant contextual details as regards HRA usage, experiences of HRA practitioners and cases of successful HRA implementation. They also reflect the ideas of actors involved in HRA more directly and have been published with less delay than the academic articles, thus potentially mirroring important HRA developments before the topic appeared in traditional academic journals.

4. Analysis

The analysis started with identifying content in each examined article consistent with the three main categories of the HRA-as-practice framework – HRA practices, HRA practitioners and HRA praxis. We assessed the articles to reveal what fits under these categories, namely what practices are discussed as a part of HRA, how they are enacted and what actors are involved. We also paid attention to other interesting topics and concepts covered in the articles in connection to the three main categories. The analysis revealed three broad topics, i.e. HRA technology, HRA outcomes and HRA hindrances and facilitators, which we call topics in the nomological network of HRA-as-practice. We discuss them below, after the main categories. These topics and their content are of particular interest for addressing the question of what connects the “three Ps” in a nomological network of HR-as-practice and what creates coherence in understanding of practice of HRA. In the analysis, the content of the articles was also synthesised into several subcategories under the three main and three related nomological categories. A complete overview of the categories and subcategories with examples can be found in Table 1 .

The results regarding the number of articles discussing the basic categories – HRA practices, HRA practitioners and HRA praxis – and additional related topics – HRA technology, HRA outcomes, and HRA hindrances and facilitators – are illustrated in Table 2 . It clearly shows that all articles analysed in one way or another address practices involved in HRA and HRA outcomes. However, the findings show that not all reviewed papers deal with HRA practitioners, HRA technology, and HRA hindrances and facilitators. Interestingly, even in the practitioner-oriented group, not all papers address these categories. It often occurs in either technical papers that focus on providing a certain statistical method for HRA or publications of a promotional character that treat companies as competing actors in the market.

HRA praxis was found to be the category addressed least in the reviewed literature. Only 44 out of the 100 articles addressed the question of how HRA practices are enacted. This number is even lower for articles within the academic group, where only 19 out of 53 articles address HRA praxis.

4.1 HRA practices

In the analysis, we aimed for a wide coverage of possible HRA practices discussed in the literature with a focus on general activities, something that is done by practitioners. We could identify several sub-practices that are seen to construct HRA. To categorise the content of the articles as a HRA sub-practice, we were looking for what is done in the organisations when they say to be involved in HRA, e.g. activities such as data collection and extraction, producing different types of analyses and reporting of results.

The multiple HRA sub-practices extracted from the analysed literature have been synthesised into four separate but related groups. The first group includes HRA practices linked to data usage , such as data management and governance. The category includes practices connected to both HR and other business data and data from external sources, such as market or industry data (e.g. Jacobus, 2015 ; Hamilton and Sodeman, 2020 ). Practices of constructing and following different measures, also called metrics or indicators, that might be relevant for HR and business strategy are included (e.g. Brown, 2020 ; Buttner and Tullar, 2018 ).

The second group includes HRA sub-practices linked to data analysis . The examined literature suggests the application of different statistical analyses at different levels of sophistication, distinguishing between descriptive (e.g. Jones and Sturtevant, 2016 ), predictive (e.g. Brandt and Herzberg, 2020 ), occasionally prescriptive (e.g. Rasmussen and Ulrich, 2015 ) and even autonomous analytics, such as in the context of autonomous algorithms (e.g. Gal et al. , 2020 ). Much attention was found to be paid to the practice of prediction: predicting valuable HR and organisational outcomes, such as employee retention or individual and organisational performance (e.g. Zuo and Zhao, 2021 ; Speer, 2021 ).

The third group includes practices related to producing data-based insights . Insight generation is mentioned by almost all reviewed articles as the central practice of HRA (e.g. Ames, 2014 ; Dahlbom et al. , 2020 ). Insight generation is seen to include visualisation (e.g. Andersen, 2017 ), storytelling (e.g. Welbourne, 2015 ) and communication of results produced by data analysis (e.g. Lipkin, 2015 ). It is these practices that are argued to be of great importance for successful HRA users' buy-in.

Finally, the fourth group includes HRA practices of decision support . Since improved HR and business decisions are assumed to be the goal of HRA, most of the analysed publications discuss sub-practices that pay particular attention to evidence-based (e.g. Hirsch et al. , 2015 ), user-tailored (e.g. DiClaudio, 2019 ), action-oriented (e.g. Jörden et al. , 2022 ) and often strategy-driven (e.g. Minbaeva, 2018 ) decisions.

4.1.1 HRA practitioners

The analysis revealed two broader groups of HRA practitioners: HRA producers and HRA users. HRA producers are the practitioners who are directly involved in the everyday activities of producing HRA, such as managing and collecting data, producing analyses, visualising results and communicating insights. HRA consumers are the practitioners who use HRA results as a basis for decision-making. HRA producers and HRA users represent both individual and collective actors, e.g. various individual professionals, groups, teams, departments, organisations and even the whole HR profession.

The most discussed HRA producers are HRA teams and their members, sometimes also called HR analysts . Such groups often include specialists from different functional and organisational areas: HR, IT and data science. There is no consensus regarding the exact organisational position where such teams are placed; placement both inside and outside HR departments is possible (e.g. Peeters et al. , 2020 ; Van den Heuvel and Bondarouk, 2017 ). Together with discussing the organisational belonging of HRA teams, the analysed articles also focus on the competences, knowledge and skills of team members and how these connect to the different areas linked to HRA. The role of an HR analyst, for example, is still in development, but several articles discuss the required competences, which are said to include technical and data knowledge, ability regarding statistical analysis, visualisation and communication and business and HR knowledge (e.g. Kryscynski et al. , 2018 ; McIver et al. , 2018 ; Minbaeva, 2018 ; Van der Togt and Rasmussen, 2017 ; Feinzig, 2015 ). A competency model already exists for the emerging role of HR analysts ( McCartney et al. , 2020 ). “HR analyst” is the label most frequently used to depict HRA producers. However, Gal et al. (2020) suggest another title, that of “algorithmists” or auditors of algorithms, named after the algorithmic technology they are supposed to apply in their HRA work.

Other categories of HRA practitioners as producers are also discussed in some of the analysed articles, such as IT and management consultants, researchers and academics and external experts. These types of practitioners, though represented by external actors, have direct influence on HRA making within organisations. IT and management consultants, for instance, commodify and sell similar technical solutions accompanied by business models and processes to several organisations, popularising HRA and making it a HR “best practice” ( Angrave et al. , 2016 ). Researchers and academics, when involved in organisational HRA projects, might directly contribute with their theoretical knowledge and rigorous social science research methods to complement and modify different HRA practices and their enactment ( Simón and Ferreiro, 2018 ). Similarly, external experts might influence HRA activities in organisations by using their expert knowledge in, e.g. the artificial intelligence area, legal and ethical requirements and diversity and inclusion questions ( Hamilton and Davison, 2022 ).

Not surprisingly, however, the most common category of HRA practitioners discussed in the assessed literature is HR departments and HR professionals. Interestingly, HR departments and HR professionals such as HR managers, HR business partners and HR specialists are considered both producers and consumers of HRA. Many articles suggest that HR professionals, especially HR managers responsible for HR decisions, are important users of HRA (e.g. Levenson, 2018 ; Nicolaescu et.al ., 2020 ; Pessach et al. , 2020 ). Some articles argue that HR professionals are the ones who also should produce HRA (e.g. Boudreau and Cascio, 2017 ; Howes, 2014 ; Vargas et al. , 2018 ). Other publications do not support this idea, arguing that traditional HR professionals and the whole HR profession generally lack the appropriate analytical skills and business acumen, which makes them incapable of producing HRA ( Angrave et al. , 2016 ).

Another group of HRA users who also attract a good deal of attention, apart from HR professionals, are top managers, including CEOs ( Shet et al. , 2021 ), line managers ( Nicolaescu et al ., 2020 ) and other types of managers and business professionals ( Barrette, 2015 ).

As a complement to the analysis, it is important to note that some of the analysed articles address broader groups of actors that have some connection to the general topic of HRA, including external actors or stakeholders such as the general public, key opinion leaders, customers, suppliers and regulatory agencies (e.g. Hamilton and Sodeman, 2020 ; Belizón and Kieran, 2022 ). Although these actors are seen as only indirectly involved in the enactment of HRA practices in organisational settings, they are important for understanding the topic of HRA, especially from a multi-level institutional perspective. For instance, such institutional actors are involved in forming general public opinion and building legitimacy of HRA practices, influencing other actors, e.g. organisations, organisational leaders and HR professionals, in their decisions regarding HRA usage ( Belizón and Kieran, 2022 ).

Employees are another interesting group mentioned in some of the articles ( Khan and Tang, 2016 ; Giermindl et al. , 2022 ). While their role is certainly worth considering as an important aspect, especially in connection to the ethical and legal requirements regarding HR data ownership, the analysed literature is still limited in addressing the employee perspective and the employees' role in the enactment of HRA practices.

4.1.2 HRA praxis

It has been more difficult to identify HRA praxis in the reviewed articles in comparison to the more stable and defined concepts of HRA practices and HRA practitioners. This might partly be explained by the elusive character of praxis, grounded as it is in actual activities, which is, thus, challenging to capture in the text, especially in non-empirical articles. We therefore based our analysis on the definition of praxis as actually situated activities when practices are enacted in context. In other words, from the assessed literature, we attempted to extract what exactly happens in organisations when HRA practices are “done” by practitioners and how abstract practices of data usage, analysis, insight generation and decision support are unfolded in the context. We further attributed such individual situated activities to a broader category in order to provide a general picture of HRA praxis. We, however, acknowledge that such operationalisation might inevitably limit the scope of multiple unique manifestations of HRA praxis that potentially exist in real life. We assume, although, that it is reasonable due to the nature of this study.

Based on this assumption, we attributed the HRA praxis described in the analysed literature either to a process of multiple steps or a particular mechanism that HRA practitioners use to enact HRA practices.

As the analysed literature indicates, a detailed process is often described as a set of logical and sequential steps that usually begin with question formulation and then move on to extracting or collecting data, building models and measures, conducting analysis, dissimilating results, acting on results and evaluating actions (e.g. Garvin, 2013 ; Green, 2017 ; McIver et al. , 2018 ; McCartney et al. , 2020 ). The reviewed articles often contextualise these steps and provide rich descriptions of how, for example, questions for HRA are or should be formulated, analysis conducted and results acted upon. These steps are often seen to intertwine with HRA practices, which point to the interconnection of the two elements. Indeed, an abstract HRA practice, for example, data usage, is translated into praxis by its enactment in a process step of extracting relevant data from the database for a given question at hand.

To exemplify a possible HRA process in a typical firm, Hamilton and Sodeman (2020) , for instance, illustrate several steps that happen in sequence: understanding firms value chain, determining significant questions and locations of data, coordinating with stakeholders, analysing data, screening for ethical concerns, making assessments for changes together with stakeholders, and, finally, implementing change together with line managers. Another example of possible HRA praxis in the form of a process is provided by McIver et al. (2018) , who describe five iterative steps: prioritise issues with the greatest potential for organisational outcomes, decide on either a data-driven or theory-driven approach, prepare and validate data, apply multiple methods of analysis and finally transform insights into actions.

HRA praxis has also been described within the literature as comprising different mechanisms for the enactment of HRA practices, such as customisation ( Jörden et al. , 2022 ), alignment to decision makers' perceptions of business reality ( Ellmer and Reichel, 2021 ), building of relationships and networks ( Collins, 2015 ), establishment of HRA’s credibility and legitimacy ( Hirsch et al. , 2015 ), exercise of strategic commitment ( Belizón and Kieran, 2022 ), demonstration of ethical ( Gal et al. , 2020 ) and legal compliance ( Hamilton and Davison, 2022 ) and encouragement of employee involvement and protection of their benefits ( Lipkin, 2015 ).

Ellmer and Reichel (2021) , for example, describe how HRA practitioners produce HRA outputs by aligning to the decision-makers’ perception of business reality. Such alignment to the final users' needs includes speaking the language of numbers, customising dashboards and boundary spanning. Thus, in this case, HRA practices are enacted through using certain numbers, particularly financial indicators, which is a common language for decision-making managers, adapting figures and diagrams for the visualisation of HRA results, and establishing relationships across diverse functional departments. Another example of HRA mechanisms is suggested by Belizón and Kieran (2022) , who argue that HRA enactment happens through the legitimacy establishment process, where strategic commitment, data infrastructure decisions and focus on HRA projects explain how HRA unfolds in practice. In this case, HRA praxis is made evident via HR practitioners' commitment to HR and business strategy, decisions on HRA data storage, whether inside HR function or as part of a companywide data warehouse, and focus on small-scaled HRA projects.

We again acknowledge that it is naturally impossible to identify all mechanisms that might be used by HRA practitioners to enact HRA practices in reality, as they are context-dependent and individual in every situation. Thus, the mechanisms extracted from the analysed literature represent only a few examples of HRA praxis described in the articles. Presenting them, however, provides an indication of how HRA practices are enacted in real life. For example, it is feasible to assume that, e.g. the practice of producing data-based insights can be enacted by the mechanism of aligning to the final users' needs. HRA producers can engage in meetings and talks with their HRA users, in this case, business managers. This allows them to better understand their managers' needs and produce insights accordingly.

4.1.3 Topics in the nomological network of HRA-as-practice

Along with identifying what practices constitute HRA and how and by whom they are enacted, this study is also interested in understanding how these three elements are connected in a coherent model and what topics exist around them in a nomological network. The analysis of the reviewed articles has revealed three topics that are widely discussed in the assessed literature. We have labelled them: HRA technology, HRM outcomes and HRA hindrances and facilitators. These categories and their content have a very clear connection to HRA practices, their implementation and their development, but they do not fall directly under any of the three main categories in Björkman et al. ’s (2014) model. We see it as reasonable and natural to bring forward these entities and conditions as clear candidates for the topics in the nomological network of HRA-as-practice. Mapping these topics and their content in the network of HRA-as-practice creates coherence between the main categories, helping to understand the practice of HRA holistically.

Accordingly, HRA technology is discussed in almost all reviewed articles, which is not surprising because the phenomenon of HRA is often linked to technology and is enabled by it. The depth of the discussions regarding HRA technology varies, however. Some articles mention technology in general terms (e.g. Vargas et al ., 2018 ; Karwehl and Kauffeld, 2021 ; Andersen, 2017 ), and some focus on one type of technology, such as artificial intelligence (e.g. Gal et al. , 2020 ; Roberts, 2017 ).

We divided HRA technology into three subcategories: general technology, HRA tools and HRA techniques. Articles dealing with general technology discuss topics of automation ( Van den Heuvel and Bondarouk, 2017 ), computerisation ( Murphy, 2016 ), cloud technology ( Feinzig, 2015 ), social media ( Leonardi and Contractor, 2018 ), big data ( Wang and Cotton, 2018 ), robotics ( Jones, 2015 ), artificial intelligence ( Hamilton and Davison, 2022 ), algorithms ( Gal et al. , 2020 ), facial recognition ( Hamilton and Sodeman, 2020 ), as well as the internet of things, biometric technology, sensors and wearables ( Holwerda, 2021 ). HRA tools include data storage and management tools, such as different organisational information systems, databases and data warehouses, with a particular focus on HR information systems as an important source of HR data (e.g. Dahlbom et al. , 2020 ; McCartney et al. , 2020 ). Another example covers tools that can carry out different data analyses or perform statistical calculations, such as Excel, SPSS, R, Stata or Python ( King, 2016 ; Ryan, 2021 ), and those examining the reporting and visualisation tools, such as dashboards and PowerPoint ( Buttner and Tullar, 2018 ; Welbourne, 2014 ). Lunsford and Phillips (2018) identify more than 300 different HRA tools and provide a detailed list of the most popular tools used by a broad range of organisations. The articles dealing with HRA techniques are focused on carrying out different statistical descriptive, predictive and prescriptive analyses, such as benchmarking ( Jones, 2015 ), data mining ( Rombaut and Guerry, 2018 ), sentiment analyses ( Gelbard et al. , 2018 ), machine learning ( Yuan et al. , 2021 ) and mathematical modelling ( Pessach et al. , 2020 ). There are thus clear links from general HRA technology, tools and techniques both to HRA practices, HRA praxis and also to HRA practitioners.

The next topic that all the articles address is HRA outcomes . We divided them into two broad groups: business benefits and HR-related outcomes. Articles dealing with business benefits focus on issues, such as improved firm performance ( Larsson and Edwards, 2022 ), revenue and ROI ( Holwerda, 2021 ), time and cost savings ( Hickman et al. , 2021 ), effectiveness ( Levenson, 2018 ), efficiency ( Zuo and Zhao, 2021 ), competitive advantage ( DiClaudio, 2019 ), increased productivity ( Lal, 2015 ), reduced uncertainty ( Frederiksen, 2017 ), facilitation of strategic change ( Hamilton and Sodeman, 2020 ) and effective strategy execution ( Levenson, 2018 ). Articles dealing with HR-related outcomes examine phenomena such as HR impact and strategic influence ( King, 2016 ), operational effectiveness of HR function ( Walford-Wright and Scott-Jackson, 2018 ), improved HR processes, such as recruitment ( Staney, 2014 ) and assessment ( Lam and Hawkes, 2017 ), employee learning ( Hicks, 2018 ), credibility and the professional legitimacy of HR ( Belizón and Kieran, 2022 ), increased individual job performance of HR professionals ( Kryscynski et al. , 2018 ), accuracy, fairness and employee commitment ( Sharma and Sharma, 2017 ), a just workplace ( Hamilton and Davison, 2022 ), and effective HRM ( Hamilton and Davison, 2022 ). One of the HRA outcomes that is commonly discussed in both groups is an improved decision-making process and better overall decisions, either business- or HR-related. This is the most frequently mentioned outcome in the reviewed literature. Better decisions are decisions that are data- and evidence-based, objective, strategic and effective (e.g. Boudreau and Cascio, 2017 ; Lunsford and Phillips, 2018 ).

An interesting observation is that all articles, in one way or another, sometimes with conditions, mention the positive outcomes of HRA, either as potential or as actually achieved. The only exception is Jörden et al. (2022), who suggest HRA has a negative impact on the HR profession because of the differences in identities and logics between managers and HRA practitioners. In sum, different HRA outcomes can clearly be linked to all main categories of the HRA-as-practice model and particularly to the idea that HRA-as-practice is a continuously evolving entity.

The final topic that is revealed from our review is HRA hindrances and facilitators . It would have been possible to discuss these two groups separately. For the purposes of this paper, however, we chose to join them together in one category because not only are they opposite sides of the same factor, but often the lack of a facilitating factor constitutes an actual hindrance. We attribute HRA hindrances and facilitators to the following subgroups: individual, technological, organisational and environmental. Individual factors related to HRA practitioners include the display (or otherwise) of different skills such as analytical and statistical ( Diclaudio, 2019 ), HR professional ( Jones, 2014 ), business knowledge and understanding ( Dahlbom et al. , 2020 ), as well as the ability to communicate ( Welbourne, 2015 ) and build relationships ( Lam and Hawkes, 2017 ). HRA users' buy-in and trust ( Lam and Hawkes, 2017 ), employees' buy-in ( Lipkin, 2015 ) and attitudes and mindsets ( Rasmussen and Ulrich, 2015 ) are also named among individual HRA hindrances and facilitators. Technological factors mentioned in the literature are either linked to data availability and quality or infrastructure and IT systems (e.g. Dahlbom et al. , 2020 ; Leonardi and Contractor, 2018 ). Organisational factors include the “right” organisational structure ( Angrave et al. , 2016 ) and analytical culture ( Ellmer and Reichel, 2021 ), resource allocation ( Simón and Ferreiro, 2018 ), operational processes ( Howes, 2014 ) and leadership support ( Hamilton and Sodeman, 2020 ). Environmental factors mentioned in the literature are privacy ( Gelbard et al. , 2018 ), ethical and legal concerns ( Hamilton and Davison, 2022 ) and the gap between academia and industry ( Rombaut and Guerry, 2018 ). The content of the topic HRA hindrances and facilitators is clearly linked to the activities of “doing” HRA. They constitute the contextual basis for action. Hindrances and facilitators are also clearly linked to the main practices of HRA and the conditions that are assumed by them. Finally, some of the features are also related to the individual characteristics of the practitioners, indicating that they might theoretically serve to integrate the main categories of our practice model, helping to constitute HRA-as-practice.

5. Discussion and future research

The analysis section has mirrored the content of the reviewed articles and constructed the current HRA-as-practice as depicted in Figure 2 . This frame and underlying theory for analysis imply, for analytical purposes, a possible separation between content belonging to the main categories describing HRA practices, HRA praxis and HRA practitioners. But it is clear from our analysis that there are many juxtapositions between them. For example, the HRA practice of data analysis is also a sequential step in a process that depicts HRA praxis. Another example is that the HRA practice of insight generation is revealed only when a HR analyst uses her analytical and communication skills in a certain visualisation activity, which is part of the HRA praxis of aligning to the final users' needs. Such observations are in line with the initial theoretical standpoint about the inseparability of practices, praxis and practitioners ( Jarzabkowski et al. , 2016 ).

According to the departure point for our analysis, the three major elements – HRA practices, HRA practitioners and HRA praxis – are inseparable in real life and thus together create a coherent whole of the practice of HRA. Our analysis has also revealed three closely related topics: HRA technology, HRA outcomes and HRA hindrances and facilitators, which are clearly linked to HRA-as-practice as a whole. These topics, together with the concept of HRA-as-practice represent the nomological network and enhance understanding of the underlying structure of the HRA field. Although HRA technology, outcomes and hindrances and facilitators were previously widely discussed in the existing literature and even categorised as HRA enablers and moderators (e.g. Marler and Boudreau, 2017 ), we tried to compile them in a nomological network of HRA, linking them to the enacted HRA practices. We have also expanded the existing categorisation of these topics by adding more recent and ample findings to their content. We suggest that these topics influence and are influenced by the combined concept of HRA-as-practice. They might also actualise the intersections between the main elements of the model and its components. Although our findings clearly show the importance of the revealed topics for HRA practices enactment, the more precise effects and relationships between them and the main categories of the HRA-as-practice framework are to be discussed in future empirical and theoretical investigations. Tentatively though, we have placed the three related topics outside the framework of HRA-as-practice in a nomological network where they are mostly illustrative for how they might influence and be influenced by the “inseparable” practice of HRA.

The topic of HRA technology covers different tools and techniques, such as HR and other organisational information systems, software for data collection, analysis and visualisation. It is found to be closely related to the practice of HRA. The proximity of technology to the concept of practice is widely discussed in the literature. Björkman et al. (2014) , for instance, in their original model of HRM-as-practice place tools including, presumably, HRM technology under the category of HRM practices. Our analysis shows that technology plays rather a different role than just simply constituting one or several HRA practices, as we understand them as abstract ideas of what is included in HRA. Technology in our suggested model has relationship not only to HRA practices, but rather actualises all constituent elements of the HRA-as-practice concept, including praxis and practitioners, by enabling abstract practices to be enacted by HRA practitioners. For instance, enactment of data analysis practice requires technology in the form of computerisation (general technology) and the application of some statistical analyses, such as regression analysis (HRA technique), using some statistical tool for data analysis, such as Excel or SPSS (HRA tool). The availability of certain HRA technology can also influence the practice of HRA with all its constituent parts: what HRA practices can be chosen, how they can be enacted, and by which practitioner. For instance, the availability of an integrated database storing data on an individual level provides the possibility for predictive analyses enacted in a set of sequential steps by a HRA practitioner with statistical skills. Conversely, HRA technology can be, in its turn, influenced by the practice of HRA. Namely, the availability of the HRA team with mixed competences, high organisational legitimacy and strategy-driven assignments at hand might influence the choice of technology to be used. Understanding HRA technology as actualising HRA practice and as influencing and being influenced by it is also in line with the idea of technology-in-practice proposed by Orlikowski (2000) , where she argues that technology is not just an artefact but manifests itself only when it is used in practice, thus converting abstract ideas of practices into evident praxis in a given situation.

HRA outcomes are another important topic in the nomological network of HRA-as-practice. As with HRA technology, HRA outcomes might also actualise the intersections between the main categories of the framework. We suggest that HRA practitioners, both aggregate and individual HRA producers and HRA consumers, are involved in the enactment of particular HRA practices depending on the potential outcomes they are seeking to achieve, thus making HRA outcomes an important component of joining practitioners, praxis and practices together. For example, a HR analyst (HRA producer) is guided by improved decision-making (HR-related outcome) when she is involved in relationship-building activities (HRA praxis) for generating data-driven insights (HRA practice). Alternatively, a line manager (HRA user) is guided by time and cost savings (business benefit) when engaging in the exercise of strategic commitment (HRA praxis) for the HRA practice of making evidence-based decisions. We also see that HRA outcomes influence and are influenced by the practice of HRA. For instance, the expected HRA outcome of improved HR reporting influences the choice of HRA practices such as visualisation of existing personnel records enacted via customisation for the line manager’s needs. On the other hand, the need for action-oriented decision support enacted via exercising of strategic commitment by the HR director might influence the choice of an expected HRA outcome, such as effective strategy execution. We also see that depending on what role HRA practitioners play, it impacts what HRA practices they draw upon and how they enact those, e.g. HR producers, such as HRA analytical teams and individual analysts, who are guided by different HRA outcomes and thus draw upon HRA practices involving data governance, statistical analyses and the generation of data insights, while HRA consumers who use such results are guided by other potential HRA outcomes and are mostly involved in the HRA practices of making evidence-based decisions.

And, finally, the actualisation of HRA-as-practice depends on HRA hindrances and facilitators. One example is that a HR analyst (HRA practitioner) involved in HR data analysis (HRA practice) by engaging in the process of sequential steps, from question formulation to dissemination of results (HRA praxis), might be facilitated by an analytical organisational culture but hindered by a lack of competence of various kinds. In line with Björkman et al. (2014) , we also assume that depending on who HRA practitioners are, it might influence how they enact HRA practices in a context and what hinders and facilitates their activities. HR analysts with a statistical background might enact practices involving sophisticated predictive data analysis in a contextual process of interrelated steps, unlike a more traditional HR practitioner, who might be inclined towards the visualisation of descriptive HR-related data through the mechanism of alignment to the final user’s needs. The praxis of these different practitioners is also potentially hindered and facilitated by different factors, e.g. specialists in statistics might be hindered by a lack of communicational skills and HR knowledge, while more traditional HR practitioners experience a lack of competence when it comes to technological and analytical skills. We suggest that this question might be an interesting and fruitful area for further empirical research.

Overall evidence from the analysed articles supports the practice perspective by clearly indicating that HRA practices have meaning only when they are implemented by practitioners. This emphasises the importance of the context in which HRA practices are enacted by practitioners. However, our study clearly shows that the context of HRA is not much elaborated in the current HRA literature. Only a few articles provide information that can be used to understand HRA praxis, namely, how HRA practices are enacted in a given context. While all the reviewed articles address HRA practices, and most of them also mention HRA practitioners, HRA praxis is only discussed in less than half of the studies. This result might be seen as a sign that the implementation of HRA is lagging in comparison to the creation of general ideas on the practices that should form the basis for any actual work. In line with the previous research (e.g. Margherita, 2022 ; Marler and Boudreau, 2017 ), we found a low number of empirical papers in our material, especially qualitative papers, with most of the studies covering conceptual research. It also goes hand in hand because studying HRA praxis in its context requires the application of qualitative methodology, such as observations and interviews. To understand how HRA practices are enacted by HRA practitioners and what contextual factors are at play, researchers must closely observe what practitioners do, say and how they interact with the environment, other actors and other things. Beneficial for HRA praxis studies would also be longitudinal approaches since the practice is under development and currently being implemented by organisations (e.g. Belizón and Kieran, 2022 ), and HRA praxis manifests itself in a certain contextual process of sequential steps and mechanisms, such as, for example, alignment to users' needs, which is naturally processual.

The important role of context is also widely supported in the broader HRM literature, suggesting a contextual approach to HRM ( Paauwe and Farndale, 2017 ). The results of this study, however, also reveal the lack of macro-contextual considerations in the existing literature, with only a few articles covering either geographical or industrial contexts, such as, for instance, the public sector in different countries. The shortage of contextual approaches to the practice of HRA is evident in the limited discussions about multiple factors influencing organisations, their HRM work in general, and consequently the shaping of HRA practices. For instance, legal requirements regarding data protection and ownership and labour union involvement might influence what HRA practices are implemented in different countries and how they are enacted by the practitioners ( Hamilton and Davison, 2022 ). The contextual macro-level praxis of HRA might also be impacted by different cultural norms; for example, the application of HRA practices in different areas, such as employee control, individual financial performance, or organisational health and wellbeing, might be more or less congruent with a certain national culture. Additionally, the lack of studies covering HRA in the public sector opens the possibility for future research to understand if HRA practices and their contextual enactment vary in business firms versus public sector organisations.

Lastly, studying HRA-as-practice in a given context is seen to have the advantages of close cooperation with HRA practitioners in different types of organisations. We strongly believe that our practice-based approach to HRA, possibly combined with some participative form of research, e.g. the so-called “engaged scholarship” ( Van de Ven, 2007 ), might generate a deeper understanding of HRA practical aspects, e.g. what activities are prioritised by HRA practitioners and why, while at the same time generating value for the practitioners in their everyday work of how different HRA practices and their enactment can solve practical problems they address. Our proposed model for HRA-as-practice benefits future research by providing a possible guideline for empirical investigations, namely, suggesting areas to cover and their potential content: HRA practices, HRA practitioners, HRA praxis and several entities actualising intersections between them, HRA technology, HRA outcomes and HRA hindrances and facilitators.

6. Conclusion

This study conceptualises HRA from a practice-based perspective by identifying what practices are included in HRA, by whom and how they are enacted and what connects them in a coherent model of HRA-as-practice. Moreover, it compiles the nomological network of HRA-as-practice, revealing what factors exist in proximity to the practice of HRA and how they influence and are influenced by it.

Summarising the results of the analysis, we suggest that HRA involves four groups of HRA practices linked to: data usage, data analysis, data-based insights and decision support. Regarding HRA practitioners, a general conclusion is that HR professionals are seen from two perspectives. They are viewed either as producers or consumers of HRA or as both. The analysis shows that most HRA practitioners are seen to be members of HRA teams, whose composition often includes non-traditional HR professionals such as data analysts and specialists in IT, statistical analysis, visualisation and communication. Findings suggest that, based on the nature of the HRA practices and competencies needed to enact them, the role of traditional HR professionals as a relevant category of HRA producers might be questioned. The study also shows that the issue of HRA praxis is the least addressed in the reviewed literature. In our analysis, HRA praxis is attributable to either contextualised processes addressing relevant problems or to certain mechanisms that HRA practitioners use to enact HRA practices.

Based on our results, we suggest that HRA practices, HRA practitioners and HRA praxis are closely interrelated and intersect. Together, they form the practice of HRA actualised by HRA technology, HRA outcomes and HRA hindrances and facilitators that influence and are influenced by it. HRA is, thus, a bundle of four types of practices, associated with data usage, data analysis, data-based insights and decision support, enacted by HRA producers and HRA users via engaging in a process of interrelated steps driven by different contextual mechanisms and actualised by HRA technology, HRA outcomes and HRA hindrances and facilitators.

Besides the theoretical contribution of conceptualising HRA-as-practice, this study contributes to HR practical work by providing a description of HRA and enabling a deeper understanding of the HRA field and how different HRA concepts are linked together. It offers HR function and HR professionals a basic ground to evaluate HRA as a potential solution to generate business value and increase HR impact by providing a holistic model, the constituent parts of which indicate the complex and highly contextual character of HRA. The model suggests that the success of HRA depends not only on the standard HRA practices that generate value as soon as they are implemented in an organisation but rather on the complex context of how and by whom such practices are enacted and actualised. HR departments and HR professionals will benefit by taking into considerations factors such as HRA technology, HRA potential outcomes and diverse HRA hindrances and facilitators that might influence the context in which HRA practices are enacted. It might potentially facilitate relevant measures when one or several named factors seem inadequate or problematic. Depending on what potential outcomes HR practitioners expect from engaging in HRA impacts, what particular practices they should implement and develop. When relevant HRA practices are chosen, their enactment is actualised by the appropriate HRA technology but can be hindered or facilitated by several factors and conditions, which are also context-dependent and require close attention in every individual situation. This means that HRA enactment in practice is highly contextual and providing a single recipe for success is problematic. However, understanding the complex contextual character of the practice of HRA might provide a useful tool for how HR professionals can work with HRA in their own individual situations.

PRISMA flow diagram

HRA-as-practice framework

Categories and subcategories in analysis