Skip to contentSkip to navigation

How to Let 999 Flowers Die

For innovation to flourish, variation must go hand in hand with selection.

(originally published by Booz & Company)

When it comes to innovation, most executives place a high value on variation. They set up formalized systems that encourage employees to generate ideas and submit them to their superiors. Even firms without a formal mechanism frequently empower their people to experiment without fear of punishment for failure. This approach is based on the knowledge that innovation is often a bottom-up process: Managers should cultivate many promising seeds to let a thousand flowers bloom.

But variation is only half the story—in true Darwinian fashion, it doesn’t work without selection. And this is where many companies fall short. Yes, choices about which ideas are worth pursuing and which are not are continually being made, but too few companies think about and organize selection deliberately, with a clear strategy in mind. The winnowing of ideas thus becomes a subjective process, wherein political interests and personal preferences determine which projects are funded and which are terminated.

I have observed the consequences of this trend firsthand at two multinational corporations that set up elaborate systems for soliciting employee suggestions, resulting in an abundance of employee proposals and ideas. In both cases, selection occurred at two points. The first was at the middle manager level. Invariably, these managers did not select the proposals that they found most promising, but instead chose the proposals they thought their superiors would want to see. They feared that passing along a bold, risky idea that their superiors might reject would be bad for their reputation. The second point started with this biased pool. Top managers picked the proposals they liked best—typically ideas that fit their preconceived notions of what the company should and should not do. As a consequence, although these companies did not seem to suffer from a lack of variation at the ideation stage, they experienced problems with their innovation pipeline.

To avoid these pitfalls, executives need to focus on developing a process that systematically manages selection in a way that aligns with their company’s strategy. For most companies, this is uncharted territory. The following five steps can help guide their way.

1.  Enable selection to happen. The first thing that top managers need to accept is that they themselves should not decide which projects live or die.

At FremantleMedia Ltd., the London-based television production company known for such popular programs as the Idols and X Factor franchises, former CEO Tony Cohen put in place a formal process that ensured deliberate selection of new television programs. Cohen often received numerous proposals. But he resisted the temptation to select the ones he himself considered most promising. Instead, he built an internal system that would identify the best ideas, noting, “Why would I know better than anyone else in the company?”

Every year, the company organizes an event called the Fremantle Market. Senior executives from subsidiary production companies around the world gather in London. For one full day, they present their ideas for new television shows to one another, usually by showing part or all of a trial episode. The creator of each proposed production explains the setup and logic behind it and answers probing questions from colleagues about audience, costs, and the potential for spillovers into other media, such as the Internet. Because Cohen set up an internal licensing system, the different executives then have the autonomy to decide whether to license the show for their own country. Hence, ideas that attract the most interest will automatically be funded. If there is a lack of interest, the proposal will instantly be selected out.

2.  Tap into the wisdom of your crowd. It’s not just senior executives who have something to offer. Taking advantage of the insight and understanding of a wider group of employees can also lead to better decisions.

At the Intel Corporation, in the days when the company still relied heavily on the production of DRAM memory chips, the company allowed engineers to work on what it called embryonic technologies. Highly skilled engineers were given individual R&D budgets and ample autonomy to decide what to work on. When top management observed more and more of these engineers abandoning projects centered on the old DRAM chips and flocking toward a new technology called microprocessors, they realized it was time to change their strategy.

For such new, innovative products, it was impossible to compute any reliable numbers (in terms of market size, demand growth, margin, or a variant of a net present value calculation), so then CEO Andy Grove relied on the collective insights of his engineers. Although Grove later abandoned this process, his successor, Craig Barrett, partly reinstated it. His “autonomous strategy processes,” though less extreme than Grove’s original, are still in use at Intel today. Companies like Google and Pixar have also adopted versions of these autonomous processes, to great success.

3.  Objectivize the process. Various research and case examples have confirmed the risk of “escalation of commitment” in selection processes. This phenomenon occurs when decision makers hold on to a failing course of action because it provided success in the past or because someone’s reputation is tied to it or simply because they have “come this far already.” To combat escalation of commitment, companies need to objectivize the process and decouple it from individual decision makers’ personal interests and emotions.

Here Intel provides another case in point. When it was producing both DRAMs and microprocessors, it let these products compete for scarce production capacity on its manufacturing line. But the company had to ensure that decisions would be made on the basis of hard facts, rather than feelings or preferences that the engineers in charge of production may have had about one product or the other. Years earlier, top management had designed a formula called the production capacity allocation rule. Using a variety of input numbers (such as efficiency, demand growth, and margins), it would compute which product would get what amount of production capacity. When it came time to make decisions about what to produce, the engineers followed the formula to a T. Even when the outcome of the formula seemed to run counter to the company’s focus—which until that time had centered on DRAMs—Grove would urge engineers to follow the formula, and with it the objective process. When microprocessors won out, it was because the data supported it.

4.  Let the evidence match the investment. Data also plays a key role in the next step. Executives often rely on just one or two selection moments. But the most successful innovators view selection as an ongoing process. As a project progresses and begins to demand increased investment, more and more data becomes available. The information revealed at one decision point should guide the next.

Consider the case of the Sadler’s Wells dance group, which operates three theaters in central London. It has an explicit mission to be the center of innovation in dance. It starts out by scouting a large variety of dancers whose work might be suitable for its theaters. It then invites a limited number of these artists to come together to develop rough ideas, in an informal way, for potential new productions. Sadler’s Wells provides studio space and a small budget to those collaborating artists who come up with a concrete and innovative idea, in order to test it. If the various people involved—artists, producers, and theater managers—believe that it has strong potential after viewing the raw idea in action in the studio, the company adds more investment to develop it into a show. Subsequently, if the show’s scale permits, it will premiere in the group’s smallest theater. If it becomes a box office success, organizers will schedule it later for a longer period in the main theater.

5.  Give them a box. Stanford Graduate School of Business professor Robert Burgelman has written that “successful firms are characterized by maintaining bottom-up-driven internal experimentation and selection processes while simultaneously maintaining top-driven strategic intent.” In other words, the variation/selection process works only if it takes place within the boundaries of a clear and explicit strategic direction for the company—as established by top management.

Innovation experts often proclaim that people need to “think outside the box” and “get out of their comfort zone.” However, what Burgelman observed in academic studies undertaken to examine the innovation processes of high-tech companies is that people actually need a box to channel their creativ-ity. This description of a company’s general strategic direction cannot be too narrow or it will inhibit new ideas, but a general direction is needed to provide boundaries within which people can innovate. For example, at FremantleMedia, the stated mission is to produce television programs that are replicable in other countries. Employees know that country-specific pitches will be immediately selected out.

Letting a thousand flowers bloom will give your company the appearance of being innovative. But investing in variation alone is not enough. You must have a deliberate system of selection to ensure that the right ideas—those that are truly best for your company’s strategic direction and not those determined by personal preference and emotion—get the funding they need. It’s the 999 flowers you let die that enable true innovation to flourish.

Reprint No. 00203

Author profile:

  • Freek Vermeulen is an associate professor of strategy and entrepreneurship at London Business School.
Get s+b's award-winning newsletter delivered to your inbox. Sign up No, thanks
Illustration of flying birds delivering information
Get the newsletter

Sign up now to get our top insights on business strategy and management trends, delivered straight to your inbox twice a week.