Benchmarking the Benchmarks

Benchmarking HR practices and metrics is valuable; many HR leaders set objectives, budgets and process improvements on these numbers. "You can learn a tremendous amount by looking at other companies—and looking beyond your company’s internal analytics," notes Bill Ziegler, managing director and global head of recruiting at UBS. If you find "companies doing better than you are, it’s worth your while to explore why."

However, appropriately using benchmarks to make decisions requires hard work. Knowing what’s behind benchmarks and when to use them make a difference.

For instance, median cost-per-hire in 2008 was $1,125, according to the 2009 Human Capital Benchmarking Study by the Society for Human Resource Management (SHRM). Its researchers use excellent data-gathering techniques and reliable information providers. However, like most benchmark statistics, the cost-per-hire number was derived from raw data that you don’t have access to, that was obtained from a variety of organizations not revealed to you and that was based on a formula that may not match your organization’s circumstances.

Underlying Data

A benchmark statistic, whether off-the-shelf or custom-built, depends on underlying data. The data set may contain both a manufacturer with 1,000 full-time employees and its nearest competitor with similar revenues and profits that employs only 100 full-time workers and outsources many tasks.

"Many practitioners use industry and company size to help select benchmark competitors. Are you comparing yourself to the whole field? Or to more profitable organizations? Identifying apples to apples is important," observes John Dooney, SHRM’s manager of strategic research.

Building appropriate comparisons is becoming harder, not easier, as corporations sharply differentiate themselves from competitors. "Companies are different now, even in the same industry," says Jac Fitz-enz, owner of the Human Capital Source in San Jose, Calif., and founder of the Saratoga Institute. For example, variety abounds even in the formerly predictable energy industry, where products and services now range from gas pipelines to wind-powered electricity. HR professionals at companies in this industry can’t just open a benchmark report to the energy section and start making decisions.

In reading a benchmark report, a statistic can be difficult to act on when you don’t know exactly what is behind it. Benchmarking organizations usually gather data by relying on the voluntary efforts of invited HR practitioners who have varying degrees of access to systems that record data and who may not have much incentive to ensure that accurate data are submitted. The compiler of a benchmark study may audit and remove suspect data—or not.

HR executives are put in the unenviable position of trying to defend the validity of benchmark data to internal stakeholders when questions arise about where the data came from and how comparable they are to the organization’s data. Even a simple statistic such as percentage of voluntary turnover can generate questions. Consider this common formula: The percentage of voluntary turnover equals the total number of voluntary terminations in a time period divided by the total number of employees in a time period.

Comparisons across organizations break down quickly when underlying data are examined. Does the numerator—the number of voluntary terminations—include full- and part-time employees? Does the company rely more on contingent labor than a competitor, thereby skewing the comparison? Even the total number of employees can be questioned: Are you using the count at the beginning of a period? At the end? An average employee count in the period? Many HR professionals have struggled with questions of comparison for years.

Benchmark reports also raise questions about data quality. The designers of benchmark surveys may "ask vague questions like, ‘What was the organization’s turnover for 2009?’ But what does turnover mean?" asks Helen Luketic, manager for HR metrics at the British Columbia Human Resources Management Association. "How did the company create the headcount average? [Survey designers] do not describe what is considered a termination, how to average headcount or even who to include in headcount."

As part of a standards-setting initiative, SHRM is working with HR professionals to alleviate some of this data confusion—beginning with a national standard for the cost-per-hire metric.

When to Use Benchmarks

Successful users of benchmarks have clear ideas of their own organizational strategies and use benchmarks with their own goals in mind. "Know your strategy first. Then, use benchmarks to gain different points of view to hone your internal performance expectations," says Brian Kelly, president of Inform Business Impact, publisher of the Annual Human Capital Benchmarking Report.

One organization’s strategy may be to use human capital to keep manufacturing operations efficient but not to invest in innovation. Another may emphasize recruiting the best innovative engineers to produce new goods. Yet another company’s strategy may be to focus only on providing consulting services in a specific niche and outsourcing other functions.

Many providers of benchmark reports—in addition to HR thought leaders—emphasize the need to use strategy as a guidepost. As David Ulrich writes in The HR Scorecard (Harvard Business School Press, 2001), "Beware of building your measurement system on ... generic benchmarks. Instead, separate out the costs associated with HR commodities ... and the unique investments required to create HR’s strategic value in your organization."

How to Use Benchmarks

Using external HR benchmark reports to make positive management decisions can yield fruit. Some best practices:

Start with a clear question. Write out the question that needs an answer and circulate it among internal interested parties. Examples of questions might include:

Is our organization providing competitive compensation and benefits?

How much faster or slower is the ratio of compensation to revenue in our company growing vs. that of competitors?

Our company is not hiring enough high-quality engineering talent. Compared to other companies hiring similar talent, are our recruiting costs within a similar range or are we spending too little?

Don’t cherry-pick reports. Too often, a benchmark statistic is used as a burning platform to force a specific decision that someone is championing. For example, an executive may want to invest in a rewards program. Five studies may not support the decision, while another report does. Since reports vary on the same topic, you may be tempted to cherry-pick the report that defends the executive’s position. Instead, look for quality research across the industry.

Factor in strategy. Sometimes, varying from an industry benchmark is a strategic decision. For example, California fast-food retailer In-N-Out Burger offers above-average benefits and wages. It would be at odds with the organization’s strategy to reduce them to meet a benchmark. Your strategy has a talent component; if you strive to meet an industry benchmark, make sure it does not conflict with your strategy.

Insist on quality. Not all HR statistics are equal. Many reports are developed by statisticians using accepted methods of data-gathering and validation. Others may be produced by a marketing team. How can a buyer tell? Read supporting sections, often labeled "Methodology" or "Data Summary."

Note the statistic used. Averages can be expressed in several legitimate ways. The cost-per-hire benchmark in SHRM’s annual survey is a median average, not a mean average. Look in supporting materials for the statistic used and why.

Pay attention to dates. When was the survey taken? The economic climate during the time data were collected affects the results. David D’Angelo, senior HR analyst for the Federal Reserve Bank of Cleveland, notes that, in his opinion, "Low turnover during a challenging economy may merely indicate a lack of available jobs rather than a successful retention program." Benchmarks that include 2008-09 data will display dramatic changes in key variables that will be hard to use in more-robust economic climates.

Take benchmarks with a grain of salt. "Look at benchmarks as an indicator but not as a literal goal. They are a challenge to do better, but there’s also an understanding that we are in continuous improvement mode," Ziegler observes. Recognize how comparable the data are to your organization’s numbers. While government leaders have a slightly better chance comparing apples to apples, every HR leader needs to consider what organizations they benchmark against.

Lee Webster, SPHR, SHRM’s director of HR standards, expects the cost-per-hire standard to be completed and ready for use by the end of 2010. Even then, a metrics standard will be no panacea. To gain powerful insight to improve human capital decisions, HR executives will need to analyze more-transparent external benchmarks and also consider internal strategy and their organizations’ needs.

___________________________________________

The author, based in New York, is senior vice president of Hodes iQ, a talent management solutions provider, and author of Ultimate Performance (Wiley, 2007).

Web Extras

• Online sidebar: How to Evaluate Benchmark Reports
• SHRM video: John Sullivan, professor at San Francisco State University, says focusing a supervisor’s attention on people management and hiring requires the use of metrics, rewards and public scrutiny.
• SHRM toolkit: HR Metrics