David Armano over at Logic+Emotion has a great story about how Benjamin Franklin used a simple graphic device to get his message across – planting a mixture of plaster and flower seeds in the shape of the words “THIS HAS BEEN PLASTERED” to demonstrate the beneficial effects of plaster on plant growth.
The tale also shows the benefits of experimentation. Franklin could have written a learned treatise about the effects of plaster on plant growth, he had lots of data to prove it, but instead, he chose to do a simple experiment.
Many companies face a much more difficult problem in that they don’t have the data to prove pretty much anything, not to a statistical degree of accuracy. Many companies overcome this difficulty by gathering more, and more, and more data. These companies often succumb to ‘one more data point paralysis’, where the search for that last elusive data point to prove beyond a doubt… becomes a game of diminishing returns.
But it doesn’t have to be that way. More and more companies are learning the benefits of gathering just enough information to run a series of experiments. The experiments not only generate their own data, but they also allow companies to develop new skills in new areas and to learn what it takes to commercialise them.
This is how I approach pretty much all of my consulting work with clients today. It has been for the past five years or more. Gone are the days when huge spreadsheets full of over-optimistic, faulty data are used to justify almost any project. In their place are portfolios of strategic scenarios and options that get turned into a series of step-by-step experiments. Increasingly, computer-based virtual experiments. In the uncertain world we live in, learning by doing experiments is the ONLY way to formulate and implement strategy.
What do you think? Should business err on the side of data-driven caution? Or should they learn by doing lots of experiments?
Add your own comment and let’s get the conversation going.
Graham Hill
