While virtually all organizations today view measurement as critical to human capital decision making, too many are still basing important decisions on anecdote, unreliable data and unsophisticated analyses. At best, this results in missed opportunities as companies fail to identify key human capital levers that could be used to improve business results; at worst, companies are misled into taking actions that actually hurt the bottom line.

Some organizations, however, are taking advantage of improvements in information technology and analytics to more precisely measure the impact of human capital on business results. Rather than basing decisions on benchmarking alone or on correlations between human capital factors and financial results, these companies are using statistical analyses to establish causation between human capital factors and critical organizational outcomes. This requires determining not only that two factors – bonus opportunity and profitability, for example – are related, but also the extent to which each precedes and drives the other. After all, to make the argument that bonus potential should increase requires evidence that bonus opportunity drives profits.

Companies are also using statistical techniques to link perceptual and behavioral data to enhance understanding. This is critical to good decision making, since what employees and employers say – as measured through focus groups, senior leadership interviews, employee surveys and review of company policies – is often at variance with how employees and employers actually behave and drive value, as measured through individual employee records that track turnover events and promotion and rewards decisions, as well as business data that track performance measures such as customer satisfaction, growth, profit and productivity. Only companies that analyze both data sets can develop an accurate picture of which HR programs and practices – and which perceptions – are actually driving attraction, retention, rewards and performance.

To illustrate how statistical analysis can lead to better decision making – and ultimately better results – we will review three real-world examples. The companies described were able to use advanced analytics to optimize rewards, better manage their talent, or improve their attraction and retention results.

Optimizing rewards

In the first example, a specialty retailer was interested in tweaking its rewards program for store associates. In the past, the company had compared its rewards package to those offered by competitors in order to make rewards-mix decisions, but these benchmarks provided no information about what the company’s own employees most valued. This time, the company relied instead on statistical modeling to review its rewards mix. Using a technique called conjoint analysis, Mercer asked employees to evaluate different paired combinations of reward elements, through which we helped the company develop a full picture of employees’ ranked preferences.

The analysis revealed that there was a misalignment between the value of the benefits delivered and the importance of those benefits to employees. For example, medical and retirement benefits turned out to be far more important to employees than some of the rewards the company had been emphasizing, such as paid time off (see Exhibit 1). As a result, the company was able to shift its overall rewards mix to better align its benefits budget to employee preferences without actually increasing compensation – thus creating a win-win situation for the company and its employees.

Managing talent

Advanced analytics can also be used to help organizations better manage their talent. For example, a major manufacturer of consumer durables was experiencing a quality problem that had led to massive product recalls costing nearly a billion dollars. The company suspected that a pervasive managerial development program for its product managers (designed to create a healthy supply of generalists ultimately available to fill top executive positions) created “employee churn” that led to this outcome.

To validate this hypothesis, Mercer analyzed the various factors associated with the promotion of the company’s product managers to see what type of behavior the company was rewarding. The analysis looked at the relative likelihood of promotion over time and found that product managers were most likely to be promoted as they reached the two-year mark in their current positions – but that the likelihood of promotion quickly diminished as their time in the position neared three years. This meant that from a product manager’s first day on the job, he or she had to begin lobbying for a new position so as not to become stuck.

The analysis revealed that a lateral move was far better than no move in opening the door for a future promotion. A product manager who moved laterally was twice as likely to be promoted in the next two years as a manager who had stayed put.

When this information was reviewed in light of the fact that it takes three to five years for one of the company’s products to move from development through to production, it became clear that the incentives created by the development program were in conflict with the needs of the business. 

The employees who were making the decisions about how to build these products were not staying in their jobs long enough to remain accountable for the final outcome of the products they were producing. The development program was successfully creating the churn intended, but that churn was negatively affecting the business. The company moved to more selective inclusion in the development program and required a new minimum time-in-position before moves would be considered.

Attracting and retaining talent

Finally, sophisticated analytics can be a powerful tool in helping companies identify the drivers of attraction and retention. For example, Mercer worked with a regional bank experiencing high turnover that was crippling branch performance, leading to long lines, poor customer service and inaccuracies. To find out what was driving the turnover, the bank conducted exit interviews, in which employees identified insufficient pay as the main reason for their departure. However, when Mercer looked not at what employees said, but at HRIS and payroll data to see what they actually did, the analysis revealed that a 10 percent market pay adjustment reduced turnover by only half of a percentage point over the long term. Pay wasn’t, in fact, the primary driver of turnover (see Exhibit 2).

Instead, the employees who were most likely to continue working at the bank were those who had been promoted in the last year and employees who performed multiple jobs. Although branch banks offer only so many opportunities for promotion, the findings of the analysis prompted the company to devote additional resources toward development rather than focusing solely on pay. Participation in incentive programs (not the amount of the incentive) was another important driver of retention. In fact, targeted investments in development programs and expansion of incentive eligibility reduced turnover by more than 20 percent.

Supervisory stability also proved to have a sizable impact on reducing turnover. A focus on retention of supervisors and targeted communications with employees when their supervisors left the bank for another employer became additional key components of the bank’s overall retention strategy.

Applying analytics to your decision making

By taking the following steps, organizations can leverage their data to make better human capital decisions – and thus create additional value for the business:

1. Identify and secure rich data sources within the organization, including:

2. Evaluate, clean and transform the data captured to create an integrated data warehouse that can be regularly updated and leveraged for analysis.

3. Identify causal connections between human capital and business variables, using a variety of statistical techniques.

4. Use the insights gleaned from analysis to design and implement solutions. The data warehouse can then be used to monitor impact.

Conclusion

As the experiences of the companies profiled demonstrate, advanced analytics can be a powerful tool in helping organizations derive the most value from their human capital investments. While instinct and experience can be valuable, only rigorous analysis can enable complex organizations to find and manipulate the many connections that exist between their human capital programs and policies and business outcomes. In addition, by allowing organizations to identify the magnitude of the impact of various actions – something that remains nebulous when HR professionals simply “follow their guts” – an analysis can help them prioritize the different actions.