Fundamentals of HR Analytics

A Manual on Becoming HR Analytical

Fermin Diez, Mark Bussin, Venessa Lee

• Publisher: Emerald Publishing Limited
• Year: 2019
• ISBN:
  9781789739633
• Topic: Business
• Subtopic: Human Resource Management
• Index: Business

The future of the Human Resource Management profession is inextricably linked to understanding data analytics. This book provides practical, hands-on approaches to connect data to HR policies and practices to help influence overall business performance.
Building on traditional HR skillsets, the book makes understanding and engaging with data analytics possible for professionals at all levels. Leveraging key statistical and financial concepts, including ROI and people productivity, and commonly available tools such as Workday and Tableau, the authors explore key skills and tasks in an accessible and illuminating way, including:

• data-analytic thinking
• data management
• data collection
• clean-up and warehousing
• building descriptive and predictive models
• applying HR analytics skills and tools to workforce planning, recruitment, training, and turnover analysis.

Fundamentals of HR Analytics is an essential resource for aspiring, new and experienced HR professionals across a wide range of industrial contexts, as well as data analysts working on business workforce focused projects and upper level students of HR encountering data analytics in this context for the first time.

Part I: The Basics of HR Analytics

1

Basics of Finance, Statistics and Data-analytic Thinking

1.1 Learning Objectives of This Chapter

In this chapter, we will first talk about why the Human Resources (HR) profession needs to embrace analytics, and speak business language, if it is truly to get the much coveted ‘seat at the table’ which dominates much of the discussion about the importance of HR. The next sections cover the eight-step methodology to be used when addressing HR analytics problems. We will then delve into basic financial and statistical concepts needed to address HR problems in a business context.

We conclude with a case story, explaining how and why this large multinational has implemented HR analytics as part of its people strategies.

1.2 The Changing Nature of HR

As organisations seek to improve performance, the onus is on HR to build value. The well-worn phrases about ‘People are our greatest assets’ and about ‘HR being the ultimate competitive advantage’, can be true when the HR function focuses on solving business problems as opposed to people problems.

The HR profession is still struggling with this idea of having a seat at the senior management table. The answer to this question is, yes, it can, if HR can speak the same language as the rest of the senior management team. As long as HR professionals continue to talk about people turnover and employee engagement as the key metrics in HR, other senior executives will continue to think that, for HR, it is people's emotions that drive business. Senior executives speak in numbers, financial and otherwise. If HR wants to be heard, it needs to be able to put HR arguments in business language, meaning, using data that link HR decisions to business outcomes. This is the key. Otherwise, discussions between HR and the line becomes one person's ‘gut’ versus somebody else's ‘gut’. HR runs the risk of being ignored if it cannot present a credible business case. This is where HR analytics helps HR to become a credible business partner: showing up with a set of well-thought-out numbers will go a long way.

Today's HR Function is expected to provide senior leadership with more information to run the business, and also provide more personalised services to employees. These demands encapsulate the arguments for the need of HR analytics in organisations: On the one hand, the HR professional that has a handle on analytics is better positioned to answer business questions from top management (e.g. ‘Which profile of our sales force will best help us to increase sales revenue?’). On the other hand, HR analytics tools can also help deliver a better employment experience to employees (e.g. ‘Which combination of employee benefits and work–life balance programmes delivers highest staff engagement?’). As data are more readily available than ever, HR is being asked for more information, better insights, and more precise recommendations to help run their businesses. It is by understanding the underlying business issues and delivering on these requests, that the HR Function can best claim the ‘seat at the table’ it rightly deserves. Fig. 1.1 below illustrates the changing HR requests from management.

Fig. 1.1.Changing HR Requests from Management.

1.3 Why HR Analytics Now?

In 2016, Deloitte's report on Global Human Capital Trends (https://www2.deloitte.com/content/dam/Deloitte/global/Documents/HumanCapital/gx-dup-global-human-capital-trends-2016.pdf) revealed that 77% of all organisations believe people analytics (PA) is important, and 32% of the participating companies felt ready or somewhat ready for analytics – a slight increase from 24% the previous year.

According to Deloitte's Global Human Capital Trends 2017 (https://www2.deloitte.com/content/dam/Deloitte/global/Documents/About-Deloitte/central-europe/ce-global-human-capital-trends.pdf), a majority of companies (71%) who participated in the study continued to report that PA is a high priority in their organisation. However, only 9% agreed that they have good understanding of which talent dimensions drive performance in their organisations; and only 15% reported to have widely deployed talent scorecards for their line managers, evidencing a slow adoption of analytics.

Despite the collection of a wide range of data from various sources, many organisations today are not effectively leveraging their data for HR analytics. Why? Deloitte's study revealed that more than 80% of HR professionals score themselves low in their ability to analyse. This is a troubling fact in an increasingly data-driven field.

For new HR professionals, analytics will become the price of entry into the profession. For existing HR professionals, analytics is the minimum expected to be able to have strategic conversations about the impact HR has in the implementation of the business' strategy.

1.4 Types of Analysis

There are three main types of analysis that can be done with data.

1. Descriptive Analysis. Answers the question: ‘What happened?’ This type of analysis summarises raw data to understand what has occurred, based on historical data, and helps to uncover patterns that can offer insights to explain the reasons for the occurrence. This allows to learn from past behaviours and understand how they might influence future outcomes.
2. Diagnostics. Answers the question: ‘Why did it happen?’ In this type of analysis typically there are measures about the relationship between two variables, and the motivation is to go beyond ‘what happened’ to understanding ‘what was the driver or explanation for what happened?’. This knowledge can then allow us to take actions that reinforce a desired outcome or mitigate against an undesired one. For instance, job satisfaction vs retention; engagement vs profitability; culture vs turnover; ethics vs profit; job satisfaction vs customer satisfaction; etc.
3. Predictive Analysis. Answers the question: ‘What is likely to happen?’ Uses a variety of statistical techniques to determine the probable future outcome of an event, or the likelihood of a situation occurring. Note that, different than in the abovementioned diagnostic analysis, there is a clear implication on the direction of causation. A causal relationship exists where the occurrence of one event is linked to another. For example, recruiting source predicts retention; changes in LinkedIn profile predicts absenteeism; training programmes predicts sales outcomes; projected return on investment (ROI) of a new talent retention solution; forecast ROI of new compensation arrangements.

Fig. 1.2 illustrates the continuum of HR Analytics Maturity, from the least to the most powerful. The descriptive techniques are often based on anecdotes (‘Our managers say that the lack of compensation competitiveness is causing attrition on their team’), reactive checks (‘Exit interviews confirm that pay is the reason our employees are leaving’), ongoing reports (‘Our engagement scores are lower than last year’), or segment and comparison, which can be from internal data or from external benchmarking.

Fig. 1.2.HR Analytics Maturity.

(‘Our competitor is providing 80 hours of training per employee vs our current rate of 40 per employee’). In all these cases, there is an explanation of what has happened, but insufficient information about how to address it. Even in the last example where there are data apparently supporting more training, there is no information as to whether this is helping the competitor, or if it would help the company.

Diagnostics are more powerful than the ongoing reports because they are motivated to understand the primary drivers of an outcome of interest, such that when actions are taken on the important drivers the likelihood of improving outcomes is materially positive. Employee engagement studies are often a good example of diagnostic analytics. These would point us to top drivers of engagement to nudge us into prioritising our follow-up actions. The initial construct assumes that employee engagement can be influenced by several factors: top leadership effectiveness, direct manager effectiveness, career advancement opportunities, learning and development opportunities, competitive pay and benefits, relationships at work, etc. The diagnostic analytics helps to determine which are the factors that indeed have the biggest relative influence on engagement. This in turn allows HR and the leadership team to help prioritise on the few factors that, when acted upon, will have a material impact on improving engagement.

It is easy to assume that one causes the other, but it could be that another factor affects both, and just looking at these two sets of data does not consider this possibility. Thus, we say the data are correlated, even if we often ascribe directionality to these kinds of results. A clearer example of causality is the impact that providing training to line employees on quality issues can have on manufacturing cost reduction. We can say, definitively, that the training helped to reduce costs, even if it may not be exact figures.

Predictive analytics will help forecast what would happen if we do this or don't do that. How much we expect sales volume to change if we change the commission structure of the sales force? This is the type of data the business leaders are most eager to receive, but also the hardest to produce.

The three types of analysis are useful, however, so it's important to know and use them all, but with some caution! Some enthusiastic HR teams and professionals may fall into the trap of jumping straight into their data and profess about their analytics prowess, without identifying the right problems that these powerful tools should be wielded upon.

The power of analytics resides not in the sophisticated tools but on the thoughtful identification of problems to which these tools are applied.

A ‘rookie’ mistake often made by those new to analytics is to spend most of their time and energy on mining data with new, sophisticated tools. The lure of ‘fancy’ and ‘cutting-edge’ analysis can bedazzle the wisest. But if this search for answers in the data is not solving a problem that has the potential of a lift in business performance or if the insights are not well understood and believed by those that eventually need to act upon the insights derived, then the analysis by itself will have no use for the organisation.

We are therefore keen to equip readers with both the analytical tools and with the mindset, intuition and process to select problems worth solving. These are the foundations to build the capability to turn insights into the behavioural changes necessary to realise the value of HR analytics.

1.5 HR Analysts as Architects

The data analytics process is akin to a house-building analogy as we will describe later. As much as it can be broken down into distinct steps, in reality it is more iterative than linear. To be successful in harnessing the value of data, readers need to embrace – and practice – some crucial mental models in addition to technical skills, which is what makes data science an ‘art’.

In the following section we explain what we mean about mental states for successful data science:

Act like an architect more than just an analyst: Once a high-impact problem has been successfully identified, it's important to frame the ‘design of...