By annie shum | August 17, 2009
By Erik Brynjolfsson and Michael Schrage, MIT SLoan School
August 17, 2009
Thanks to technology, change has never been so easy—or so cheap
CALL IT INNOVATION on steroids. Or innovation at warp speed. Or just the innovation of rapid innovation. But the essential point remains: Technology is transforming innovation at its core, allowing companies to test new ideas at speeds—and prices—that were unimaginable even a decade ago. They can stick features on Web sites and tell within hours how customers respond. They can see results from in-store promotions, or efforts to boost process productivity, almost as quickly.
The result? Innovation initiatives that used to take months and megabucks to coordinate and launch can often be started in seconds for cents. And that makes innovation, the lifeblood of growth, more efficient and cheaper. Companies are able to get a much better idea of how their customers behave and what they want. This gives new offerings and marketing efforts a better shot at success.
Companies will also be willing to try new things, because the price of failure is so much lower. That will bring big changes for corporate culture—making it easier to challenge accepted wisdom, for instance, and forcing managers to give more employees a say in the innovation process. There will be even better payoffs for customers: Their likes and dislikes will have much more impact on companies’ decisions. In globally competitive markets, they will ultimately end up getting products and services better tailored to their needs.
Already, this powerful new capability is changing the way some of the biggest companies in the world do business, inspiring new strategies and revolutionizing the research-and-development process. “In the U.S., we do the vast majority of our concept testing online, which has created truly substantial savings in money and time,” says Joan Lewis, global consumer and market knowledge officer at Procter & Gamble Co.
Finding the Link
What does all of this experimentation look like in practice? Consider Google Inc., which runs 50 to 200 search experiments at any given time. In one case, Google asked selected users how many search results they would like to see on a single screen. More, said the users, many more. So Google ran an experiment that tripled the number of search results per screen to 30. The company found that traffic declined.
What happened? On average it took about a third of a second longer for search results to appear—a seemingly insignificant delay that nonetheless upset many of the users. The greater number of results also made it more likely that a user would click on a page that did not have the information he or she was seeking. In an environment where experimentation is this quick and efficient, many traditional practices make less economic sense. For instance, current research-and-development efforts are often driven by considerations that the company’s technicians think are important but customers really don’t care about. Mobile-phone companies, for one, had a reputation for piling on features that added more cost and complexity than value.
But it’s not just Web-based companies that are taking advantage of technology to run crucial experiments. Even retailers—who might seem to have tremendous logistical challenges implementing rapid experiments across lots of stores—can tap this new approach, thanks to the rise of sophisticated tracking systems that make measuring customer behavior more agile and less expensive. These systems—which track everything from purchases at the cash register to how products move through the supply chain—allow stores to cheaply collect terabytes of data on their customer interactions, the performance of products in the field, employee productivity and other factors. Traditionally, companies have simply rooted through all of that data to look for patterns and trends; they’ve mined their data. But some retailers are beginning to realize that they can get much better results by using their digital systems to run experiments.
Take Wal-Mart Stores Inc., which frequently runs comparative in-store experiments with signage, displays and shelf layouts to see what influences shopper decisions. Wal-Mart can test different arrangements in a number of stores for a week or so to see which approach boosts sales the most. If different sign colors or unusual merchandise juxtapositions increase—or depress—relative sales, the stores can quickly share the information and implement the same plan. That said, store managers also have the ability to make decisions at their store to meet individual customer needs.
Other companies, meanwhile, have blended digital technologies and conventional systems to test out ideas. For instance, Harrah’s Entertainment Inc. uses advanced information systems to analyze data from its Total Reward Card and field experiments throughout its casino and hotel network. Gary Loveman, the chief executive who brought the experimentation mindset to 70-year-old Harrah’s, quips, “There are two ways to get fired from Harrah’s: stealing from the company, or failing to include a proper control group in your business experiment.”
Experiments to Come
Where will all this lead? Experiments will become far more pervasive and persuasive as information technology improves and testing grows faster and cheaper. More companies—and more enterprising individuals who work in them—will recognize that experiments don’t have to be time-consuming and expensive, and they will propose more tests that exploit those economies.
Increasingly, the more innovative companies—the Googles and Harrah’s of tomorrow—will shift away from traditional research-and-development methods. Five years ago, for instance, a leadership team might have reviewed two or three “big” innovation proposals from consulting gurus; executive teams today might compare the outcomes of 50 or 60 real-world experiments to decide which ones to act upon.
All of which guarantees huge changes for corporate cultures. Challenging conventional wisdom, for one thing, becomes immeasurably faster, cheaper and easier. And there’s a subtle shift in how people view innovation. People don’t talk about running experiments months into the future—they’re into immediacy, because they see that they can test ideas right away, and the company culture starts to actively encourage speed. A provocative hypothesis proposed during a morning meeting might graduate into a full-blown experiment by day’s end.
Even if a test doesn’t produce a workable idea, there’s usually something important to be gleaned from it. “Genius is born from a thousand failures,” says Greg Linden, an entrepreneur who has been an innovator at both Amazon.com Inc. and Microsoft Corp. “In each failed test, you learn something that helps you find something that will work. Constant, continuous, ubiquitous experimentation is the most important thing.”
Giving Up Control
This new environment also has big implications for managers. Simply put, bosses must be prepared to give up some control. With testing so cheap, easy and accessible, there’s less need to ration it as they have in the past. Managers used to directing the company’s innovation efforts must give their workers the freedom to come up with ideas on their own and pursue them without lots of red tape.
Some of the best experiments come from outside the chain of command. At Amazon, for instance, innovators initially developed the company’s recommendations feature—which suggests other products customers might want—without explicit approval from higher-ups. Not only do we expect managers to solicit and welcome more ideas from lower down in the ranks, we expect that lots more people will be invited to review experiments and make changes. Customer-service and maintenance people, say, might chime in on experiments proposed by manufacturing or distribution.
That might seem hard to believe, given the turf battles that can arise over new ideas. But when experiments become more abundant—and easy and inexpensive to change—those proprietary issues won’t come up as much. Think of the difference between improving and sharing presentations back in the day of transparencies and 35mm slides versus PowerPoint presentations, which can be modified with a few clicks of a mouse. As more people get involved in experimentation, companies will also need to change their focus in education and training efforts for innovation. Instead of just getting workers to creatively interpret large volumes of data, companies will need to help them develop the skills to rapidly design provocative experiments. Passive analysis will be subordinate to active experimentation.
Another crucial development down the road will be “scaling.” Digital technology, as we have seen, allows companies to test new ideas quickly and easily. But the technology also lets companies easily scale those ideas—or implement them rapidly and cheaply throughout the whole business. We predict that as companies realize the power of this idea, they will focus on experiments that not only can be tested rapidly but also can be put into wide effect just as quickly.
The most obvious example of scaling is a Web site. Companies can test out a new feature with a quick bit of programming and see how users respond. The change can then be replicated on billions of customer screens. But that kind of scaling becomes rapidly possible in the bricks-and-mortar world, too, as more business processes become digital—supply-chain management, customer-relationship management and so forth. When a retailer identifies a better process for screening new employees, the company can embed the process in its human-resource-management software and have thousands of locations implementing the new plan the next morning.
We think the future of innovation and the future of experimentation will continue to evolve, thanks to the improving economics of digital technologies. As a result, the next decade of innovation in the global marketplace will be even more tumultuous than the last. That’s a great opportunity for innovators—and even better news for customers and consumers world-wide.