Once you have advocated, screened, and funded ideas, the next step is to engage in experimentation. To experiment is to try something new. It allows you to observe the interplay between cause and effects—i.e., it is the application of scientific methods to generate actionable knowledge. Simply put, experimentation can be considered the sum of all activities we engage in to test the feasibility and elasticity of an idea. On the feasibility side, we are normally looking at the cost, benefits, effort, resources, and risks involved in transforming the idea into a viable product and/or service. In addition to feasibility, the experimentation process will also unearth the elasticity of an idea. Idea elasticity focuses on the reach of the product and/or service. Elasticity tries to see how far you can stretch ideas, both in terms of the number of products and/or services that you can develop from them, and also the domains in which you can deploy them.

It is through the process of collecting data through the conduct of experiments that we can find support for the ideas. Data collected during the experimentation process will help us gather the necessary evidence to support decision-making. Today, there is a big movement in medicine called evidence-based medicine (EBM), which demonstrates that a move towards more scientific and data-driven decision-making can prove to be valuable, rather than purely relying on gut instincts. Dr. Dave Sackett, a pioneer in the field states “EBM is saying rather than just rely on tradition, expert opinion, wishful thinking, let’s try and find the evidence and apply it.” [1] To build a culture of experimentation one must focus on the following principles: 1) do not just discard ideas without adequate evidence, 2) do not support or move ideas ahead without adequate evidence, and 3) always look to exploit data from experiments.
While at Amazon from 1997-2002, Greg Linden prototyped a system that would make personal recommendations to customers as they checked out. Linden commented, “I heard the SVP was angry when he discovered I was pushing out a test. But, even for top executives, it was hard to block a test. Measurement is good. The only good argument against testing would be that the negative impact might be so severe that Amazon couldn’t afford it, a difficult claim to make.” [2] Linden’s experiment showed how much the customer liked the feature and it won praise – the end result is that this has become a signature design feature for Amazon, and most online web marketers have introduced a similar concept. This illustrates the value and capabilities of organizations to test incremental ideas and achieve innovation through “continual tiny experiments in such areas as business processes and customer relationships rather than a single, company-transforming idea.” [3]

Experimentation needs to be made part of every employee’s work and has to move beyond the R&D Labs. The R&D Labs have natural constraints that leave a lot to be desired in terms of experimentation. For example, most of the R&D personnel are detached from the day-to-day running of the business and hence are not the best people to experiment on the problems and solutions of interest for today. In addition, these labs are often physically secluded from the operational centers of the business. This separation leads to difficulty when you try to transport (mobilize) ideas from the lab in order to address problems that are happening on the ground. Finally, you also have a numbers issue. The number of individuals working in an R&D lab is minimal compared to employees who are involved with the day-to-day running of the business. As such, no matter how brilliant your R&D lab personnel are, you will be at a loss if you cannot find ways to tap into the 85-90% of your organization’s employees who do not work in the lab. At the Engaged Enterprise, we had a R&D lab, the Institute for Engaged Business Research (IEBR). IEBR was focused on working on applied management problems that had value propositions to our clients. We determined upfront that simply relegating experimenting and innovation to the labs was not optimal. We needed to find a way to blend the experiences of those working on the consulting side with the R&D side. In addition, we needed to find ways to take knowledge that was being generated on the consulting side (as experiments were conducted “live” while projects were being done – i.e. as we tried to install a new service or strategize with a client – we were in essence engaging in experimentation) and move these into the R&D lab. We did this by having people share their time between consulting and R&D. Also, we had our R&D folks develop a method and handbook that could be used for experimentation which explained the basics of experimentation and how to capture and store results. We also encouraged sharing of results from experimentation efforts so that others might use the results, or provide their reflections on the experiments.

The effort to move experimentation beyond the R&D labs needs to be a conscious one. Both, organizational (management) and employee level interventions need to be in place to promote this concept. Managers should not only encourage their employees to experiment with their ideas, but even go so far as making it a requirement when ideas are being developed and proposed. In addition, employees should take responsibility to engage with the experimentation process, and be aware of methods and practices for conducting experiments.

please stay tuned for my new book or send me a note via email…

[1] Sackett, Dr. Dave. (October 30, 2009). Interviewed by André Picard. Available at: http://www.theglobeandmail.com/life/health/when-we-began-we-were-almost-pariahs/article1344833/
[2] http://glinden.blogspot.com/2006/04/early-amazon-shopping-cart.html
[3] McCann, David. (March 15, 2010). Testing, Testing: The New Innovation Game. Available at: http://cfo.com/article.cfm/14482988?f=search

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.