Why You Should Experiment with Your Business

How to use business experimentation to grow your company

Innovation is the basis for how new business strategy is developed and implemented. Seeking ways to be more innovative is nothing new - senior managers have been seeking this Holy Grail for years. What is different today is the increased use of controlled experimentation, both inside companies and with external partners. Executives, including CIOs, are beginning to understand the power of a well-defined business experimentation process as a way to increase the pace of organizational learning and thereby accelerate the development of new or extended products, services and processes.

Companies such as Capital One have been able to invent entirely new business models using experimentation. Much of Wal-Mart's success over the past 30 years can be traced to its well-honed competency at business experimentation. And General Electric has publicly committed to 8 percent organic growth (a very high number for a large, mature company, translating into $US13 billion in new revenue per year, on average), which depends on an experimental approach that it calls "imagination breakthroughs". Learning how to innovate quickly and cost-effectively in the areas of services, processes and products is the core concept of business experimentation.

As CIO, you can lay the groundwork for successful business experimentation. Experiments rely on information systems. But your responsibility goes beyond lending support; experimentation can also become a part of how you run the IT department.

How Business Experimentation Works

Business experimentation is both a process and a discipline, and it's used to create systematic innovation and improvement, which in turn support organic growth. We define business experimentation as a controlled, cost-effective and iterative approach to learning about the potential success or failure of a new product, service or process.

Much of Wal-Mart's incredible success can be attributed to its willingness to experiment. Initially, the small town of Bentonville, Arkansas, Wal-Mart's headquarters since 1970, provided a well-defined business model for success in other small towns. But to expand beyond small towns, Wal-Mart had to modify its strategy in ways that would allow it to be successful in Miami, Houston, Denver and other urban settings.

For example, as part of its low pricing strategy, Wal-Mart wanted to ensure that shoplifting was kept to a minimum. Managers built a series of experiments around the use of greeters in their stores to learn what characteristics reduced shrinkage. They found that elderly greeters who personally welcomed customers kept shrinkage low. Thieves were less inclined to steal from someone who looked like their grandparents. But it took several iterations and a willingness to "fail" with some of the experiments to learn exactly what type of person would make a good greeter.

More recently, Wal-Mart has been experimenting with offering in-store financial services. The chain's approach this time is to build alliances with successful financial services companies such as SunTrust Banks. The two companies have launched a set of 45 in-store bank branches called "Wal-Mart Money Centres by SunTrust" to test the new strategy. Business experimentation has become a core competency at Wal-Mart and is deeply embedded in its executives' strategy formulation process.

Page Break

The CIO's Essential Role in Experimentation

As the CIO, you play a key role in making business experimentation a core competency in your company. The discipline of experimentation, the collection and analysis of data, the presentation of that analysis to the business in a way that encourages learning, the storage of and access to past experimentation data in a way that's useful for future experiments, and the systems that present an experiment's results for optimal learning - all these are tasks performed by information systems. You should review your IT architecture to ensure that your standards and policies enable rapid and cost-effective experiments.

But supporting business experimentation is only part of your role. CIOs are well-positioned to champion experimentation within the enterprise. You can do this by making experimentation a part of IT operations strategy.

Train your staff in the process and discipline of business experimentation. Develop skills in your staff that will enable them to identify, build, run and analyze experiments. Organize for rapid experimentation by examining and redesigning routines, organizational boundaries and incentives. Train small groups of key people to iterate rapidly and learn from each iteration. Above all, make clear that IT employees need to focus on organizational learning.

Experimental Trials

To some executives, business experimentation may feel like they're giving up control of a key part of strategy formulation. After all, if they let any part of the enterprise try out new ideas at any time, it might be difficult to rein it back in when the time comes to formulate a single business strategy. In the current business climate, however, this way of thinking is wrong. Organic growth has become an important criterion for market valuation, and the rate of innovation is a key input for the rate of organic growth in large companies. Innovation from anywhere in the company is a good thing.

From the research we have conducted with our partners at The Concours Group consultancy, we have found that the greatest benefit from business experimentation occurs when an organization is able to integrate the approach into its everyday processes. The success of experimentation depends on an organizational culture that accepts learning as a goal.

It's not an easy feat to create an environment that walks the line between so-called failed experiments - where the discipline of data collection, analysis and iteration results in learning even if the experiment itself doesn't produce a desired result - and the frivolous waste of resources, where ideas are tested in an undisciplined manner. Using an experiment as a justification for a one-off project doesn't create a culture of experimentation. That kind of approach to experiments signals that managers have to "get it right the first time", and that failure is not acceptable.

The challenge for senior managers, then, is to establish an environment in which experimentation can flourish, while at the same time building processes and controls to ensure that resources are not squandered. This is certainly a complex challenge - which is why we believe it will partially distinguish between business's winners and losers in the future. Business experimentation might be the only process that is designed to continually produce the organic growth and operational efficiency your company needs to remain competitive.

James Cash is the emeritus James E Robison Professor of Business Administration at Harvard Business School. Keri Pearlson is a research director with The Concours Group and co-author of Managing and Using Information Systems