Do You Have A “Plan B”?

Image courtesy of Google.

As the scientific community has long known, experimentation is the key to knowledge. It enables people to check their assumptions and run various “what if” tests to expand the base of what they know. In the business world, executives generally recognize the importance of experimentation, but many don’t practice it nearly as much as they should. For such executives, the recently published Getting to Plan B: Breaking Through to a Better Business Model (Harvard Business Press, 2009) should be required reading.

Written by John Mullins, an associate professor at London Business School, and Randy Komisar, a partner with the venture capital firm Kleiner Perkins Caufield & Byers in Menlo Park, California, the book makes a strong case that organizations need to experiment regularly to overhaul a business model that’s broken or to fine-tune one that needs adjusting. Indeed, many successful companies originally had very different business models from the ones they eventually adopted. Getting to Plan B is filled with such case studies, from the famous (the online payment service PayPal Inc. was conceived to sell security software for handheld devices) to the less well-known (Mumbai-based Pantaloon Retail [India] Limited, now India’s largest retailer, was founded as a manufacturer of men’s pants). The common theme of the examples is that some smart executive realized that “Plan A” wasn’t working and, through insightful experimentation, was able to devise a better “Plan B.” (For some businesses, “Plan B” was also a dead end, leading them to “Plan C” and maybe “Plan D” and so on.)

Perhaps Getting to Plan B’s greatest contribution is that Mullins and Komisar provide a powerful, high-level framework that helps managers uncover and investigate weaknesses in their business models. Specifically, the framework helps executives probe any “leaps of faith” in their assumptions with respect to their anticipated revenues, gross margins, operating model, working capital and investments. Yet although Getting to Plan B provides a top-level view to identify trouble spots in a forest, it’s less useful for managers once they’re down among the trees. That is, Mullins and Komisar don’t really provide the crucial, nitty-gritty details of how to experiment.

To some extent, that type of “how-to” information has been covered in an earlier book — Experimentation Matters: Unlocking the Potential of New Technologies for Innovation (Harvard Business School Press, 2003) by Stefan H. Thomke, a professor at the Harvard Business School. In that book, Thomke describes the use of computer simulations, rapid prototyping, combinatorial chemistry and other technologies to perform faster and cheaper experiments. But Thomke focuses mainly on product R&D, which is just a subset of the type of experimentation that companies can — and should — perform.

Indeed, Mohanbir Sawhney, a professor at Northwestern University’s Kellogg School of Management, and his colleagues have identified 12 dimensions in which companies can innovate, including brand, supply chain, organizational structure, processes and customer experiences. The problem is that, although many companies may be adept at experimentation in product R&D, they may have little idea how to run similar kinds of tests in other areas, for instance, to streamline their supply chains.

Of course, ask a group of executives if they’d like to do more experimentation in their companies and the answer will probably be yes. After all, what manager wouldn’t want to know how a 5% price increase would affect sales of a particular product line, for example, or whether a new bonus plan will actually lead to increased employee productivity? That type of knowledge is invaluable, but experimentation in a business is easier said than done.

For one thing, companies need to have a culture that encourages employees to investigate numerous “what if” questions that might eventually lead to innovation. Such is the case at Google Inc., where experimentation is an essential organizational process and mind-set, as evidenced by the company’s famous 80-20 rule: Technical employees spend 80% of their time on their assigned work and are free to use the remaining 20% to explore projects that they choose. Every day, Google conducts experiments, relying on its tens of millions of daily users as test subjects. This intense trial-and-error process lets the market decide which new ideas to pursue and which to abandon.

Indeed, one of the beauties of the Web is that it enables cheap experimentation — a characteristic that companies like Google are quick to exploit. But what about traditional brick-and-mortar businesses that are not inherently oriented toward electronic experimentation? One obstacle for those organizations is that running tests takes valuable time and resources. But that’s when you need clever experimenters, people like Dan Ariely, the noted behavioral economist and professor at Duke University. Ariely has been able to plumb the depths of human behavior in ingenious tests conducted on relatively modest budgets. For example, to investigate how people tend to overvalue the things they possess, he studied the prices that the “have not” Duke University students were willing to pay — and that the “have” students were willing to accept — for coveted tickets to a basketball game against archrival University of North Carolina.

Ariely has devised numerous experiments that companies could benefit from, but the business community hasn’t always been receptive to that work. Take, for instance, his research to study how teams might improve their decision making by avoiding various common traps like the phenomenon of “groupthink.” In one project, Ariely developed software that would investigate how, for example, the presence of an authority figure might affect the way people vote for a particular proposal. Unfortunately, Ariely couldn’t get any company interested in using the software in a real-world experiment because, according to him, organizations have placed tremendous confidence in their group decisions (especially those made by their boards and other executive teams) and they’re not particularly interested in having that belief challenged.

That kind of organizational dynamic may explain why many managers don’t experiment as much as they should. After all, why run tests that might disprove a cherished belief if you’re not willing to discard it and alter your behavior accordingly? For any managers with such a mind-set, Mullins and Komisar have some sage advice: “Our working assumption is that part or all of Plan A is wrong. By systematically testing a series of hypotheses, the savvy entrepreneur or street-smart executive identifies, through experimentation rather than impassioned persuasion, a better Plan B or, eventually, Plan Z.”