Skip to main content

J.C. Penney: The Best Case for Test and Learn


By Rafi Musher, [email protected]

As much as investors may think there is opportunity for big change in retail, Monday’s departure of J.C. Penney CEO Ron Johnson highlights the need to trust your managers in the field and use “test and learn” as the best way to gauge the potential impact of changes before rolling them out. Why take radical risk with a large pool of capital when you can test a variety of ideas with low risk and then scale up those that work? If Johnson had done this, he might have avoided the risk to investors in blowing up a model that’s wasn’t working, instead of methodically using data and analysis to fix it.

Johnson’s departure came as no surprise. Since his arrival, Penney’s sales are down 25% and shareholder value is down 50% – to the tune of $5 billion. Customers are disenfranchised, and people will lose jobs. At the root of it all was Penney launching a major overhaul of its stores without testing the concepts first.

Johnson came to Penney from Apple, a company built on the principle of creating breakthrough products that the consumer may not know they need yet, so he was clearly in his comfort zone not relying on near-term customer input and constant data feeds. But technology has a different R&D cycle and higher barriers to entry for competitors than most retail environments.

Easy to test the waters

Many retailers already know that test and learn is a more tightly controlled way to introduce new ideas. You can measure outcomes in a controlled manner, and forecast the impact across the chain. And you can test almost everything: store format, product, pricing, staffing levels, technology, and, of course, marketing and advertising.

As the Penney case illustrates, a few things can get in the way of this. First, you have the “I know it, I’m a merchant” attitude. Sure, there are people who have great taste and an innate sense of what the customer wants. Malcolm Gladwell, in his famous 2005 book Blink: The Power of Thinking Without Thinking, observes that those good at these kinds of seemingly instantaneous analyses are really making thousands of observations and doing the math in their heads. While it is great to have this innate ability, customers are becoming less homogeneous, and the data feeds of today’s retail environment are too much for even great minds to calculate. Even Einstein used to write things down that he didn’t need to memorize.

Second, you have the inclination that “we need to do something big, and do something fast.” This usually creates even bigger hurdles as everyone spends time thinking about what has to be a big bold change, with solutions too risky for anybody to implement except the CEO. It’s lousy process to have a company run entirely by gut, rather than giving data to many people, so they can use it to make more decisions with better information more frequently.

More informed decisions

The defined principles and mathematics used in test and learn can bring a lot of benefits. You analyze your stores or customer segments to identify similar behaviors and reverse engineer smaller segments, even micro segments that represent larger ones. For instance, you could look for 30 specific stores that would behave like a microcosm of 500 stores. Once you know this, you can test an idea with those 30 stores to see how it works — or doesn’t — before deciding whether to drop it, tweak it, or roll it out as is.

Mike Haaf, former CMO of Food Lion, grew $3 billion of revenue in four years in the intensely competitive southeast by using test and learn. According to Mike: “We created an environment where people were encouraged to try things out without fear of failure. You could clearly see what worked and what didn’t. Out of this came innovations that you might not have been willing to try otherwise, because it would have been too expensive if you were wrong.” By allowing more ideas to be tested, using test and learn brings out more people with more ideas simply by virtue of its reduced penalty for failure.

Perhaps the most staggering thing in the Ron Johnson commentary is that, given the chance, he would do it the same way again. As a management principle, it seems just wrong to be top down in a world where the data is far more broadly available and you need teams in the field to think, come up with ideas, and have a way to test and implement them in a low-risk setting. This is the process of innovation. People need to have the chance to take risks, mitigate risks, and learn when something is not working so that they can turn it off and learn from those mistakes.

Bruce Barkus, former CEO of RGIS (a Blackstone Company) and COO of ULTA Beauty, has worked with a relatively new theory developed by Dr. Craig Pearce called Shared Leadership. This empowers those in the field with expertise to take lead roles. “In a number of companies, I have seen this approach work very effectively. The result in each case was well above industry growth, margin expansion, and building a stronger management team that led to longer-term equity value. It is critically important to get the feedback from those doing the work every day to tap into the best thinking of what will work, and then test the concept for effectiveness.”

A common complaint today is that people aren’t bold and willing to take risks. Losing is fine; it happens. What matters is what we learn from the experience and how we adjust, train, and get better. Investors also tend to look for singular leaders with all the answers. Great leaders don’t need to have the answers, just to ensure their people have what they need to get them themselves. They also recognize that getting the answers is unlikely to be a straight journey; they expect a few failures along the way. De-risking those failures through a better use of data and feedback loop, reducing the capital risked, and the interpersonal attitude are all critical to making this happen to drive innovation. If there’s a lesson in the J.C. Penney case, it’s that relying on the data is a surer path to success than intuition.

Rafi Musher is founder and CEO of the global strategy consulting firm Stax Inc (, and is based in the firm’s New York office. For the past 20 years, he has been leading corporate management consulting engagements and working with investment firm leadership teams to identify profit opportunities and mitigate risks across industries. He can be reached at [email protected].

More Web Exclusives/Guest Commentaries

This ad will auto-close in 10 seconds