MAY15_07_10172648

Do you have an accurate sense of how your company stacks up?

Figuring out a company’s relative performance is ferociously problematic. It depends on which other companies are included in your comparison. Just change the peer group, and a laggard becomes a leader, or vice versa.

And if a company’s relative performance is in doubt, so are its goals, because the two are tightly linked. When Jack Welch was CEO of GE, he famously tasked each business with achieving number 1 or number 2 status in its industry, a goal-setting principle that echoes across the decades. But applying this benchmark effectively requires choosing a comparison group, and there are two challenges that inevitably crop up: what we call the “microscope” and “telescope” problems. That is, you get a meaningless result if you look at too small a comparison group, and an equally meaningless answer if the sample is too large and heterogeneous.

For example, what if your industry has just a few key players, and they’re perennial poor performers? Is it wise to compare yourself only with them? And what about size? Can you measure your organization’s performance against firms whose annual revenues are orders of magnitude different? And then there’s the time factor: Does being “number 1” this year offset years of poor results? What about the effect of macroeconomic ups and downs?

With questions like these in mind, we recently set about developing a benchmarking method that would give executives the ability to solve both the microscope and the telescope problems, allowing them to compare their companies against a very large sample, but on a more-or-less equal footing.

How is this possible? Our method relies on specialized regression analysis, which is designed to correct for company-specific factors that affect performance. It also is designed to assist you in identifying your company’s percentile rank among all publicly listed U.S. companies, as if yours and all the other firms were roughly the same size and in approximately the same industry.

We believe that the method, which has been published in a top peer-reviewed journal, is statistically sound. Best of all, it’s easy to use. Any company, public or private, in any industry, can access our method via a simple online tool to begin to evaluate its performance. The tool is accompanied by a detailed technical note.

The approach yields some revealing results, especially as compared with more widely used methods. We’ve plotted five randomly selected companies’ 2013 ROA rankings according to three methods: a comparison with the overall market, a comparison with industry peers, and our statistical method. The data points are percentiles, which show the proportion of companies with worse performance (so if a company is in the 90th percentile, its performance is better than that of 90% of others in the comparison group).

We’ve changed the companies’ names, but the data points are real. ABC and Acme appear to be doing quite well relative to the market: If you make a simple comparison, they’re around the 80th percentile. But once we apply our method to correct for industry, size, and long-term performance, it becomes clear that they’re little better than middle-of-the-road.

By contrast, if Gizmo compared itself with its closest peers, its leaders would probably conclude that urgent improvements are needed. But our method shows that the company is doing quite well, given its circumstances. Finally, Blackacre and Widgets do well in their industries but poorly versus the markets.

W150415_RAYNOR_DIFFERENTVIEWSv3

 

These five companies are not outliers. For the full population of U.S.-based, publicly traded companies, the average absolute difference in percentile rank between our method and the more common approaches is between 18 and 25 percentile ranks for both profitability and revenue growth. That means it would be typical for a company that considered itself to be in the top quarter of its peer group to be, based on our statistical approach, no better than average. Or for a company’s simplistic benchmarking approach to mask, in part, its excellent performance.

Think about that: You might discover — to your surprise — that your company is a standout.

And you wouldn’t be alone. To probe corporate leaders’ sense of where their companies stand, we polled 301 executives from large, U.S.-based corporations and asked them to report a recent performance figure (ROA of 5%, for example) and estimate that data point’s meaning in terms of percentile rank, taking into account their companies’ industry and size. We also plugged each data point into a version of our statistical model.

When we compared the results, there was little correlation between the executives’ percentile estimates and our statistical findings. These results closely parallel two earlier survey efforts we undertook. With more than 800 executives polled overall, we have seen little evidence that business leaders can readily assess their firms’ relative positions.

Setting corporate financial goals isn’t simply a quantitative exercise in prediction, of course, and it never will be. Goals are aspirational. Ideally they reflect a healthy tension between what a company can achieve and what its key stakeholders want it to achieve. Yet goal-setting inevitably begins with an answer to the question “How have we done so far?” Goals mean little if a company lacks a clear understanding of how the organization stacks up.

The potential downside from getting it wrong is significant: Underestimate your performance and you risk setting your sights too high, whipping a horse that’s already running as fast as it can. Overestimate your performance and you could end up settling for a lazy canter, when there’s a full gallop to be had.

About The Author

Head Stag

Joseph Doyle is an active entrepreneur and life coach with a multi million property portfolio and advertising and marketing agency boosting large international brands. Contact Joseph at www.digilab.ie