This is Phil_hendrix's Typepad Profile.
Join Typepad and start following Phil_hendrix's activity
Join Now!
Already a member? Sign In
Recent Activity
Discovery-driven Planning and Agile Market Research - An Antidote to Doubling Down Prematurely? Dave, good post. Should provoke a lot of discussion. I'd like to share some supporting perspectives and offer alternative solutions. Two academics - Ian MacMillan (Wharton) and Rita McGrath (Columbia) - have examined many of the issues you raise in their discussions of innovation, uncertainty and investment. MacMillan puts it succinctly when he recommends "spending imagination before you spend your money and... engineering the risk out of uncertain projects..." In a nutshell, the process he and McGrath advocate involves 1) creating tests to probe and reduce the uncertainty ahead of investment; 2) staging the investment tranches, contingent on intermediate outcomes; 3) postponing investment until (some of the) uncertainties are resolved. Useful sources on their perspectives include notes from a conference last month at Wharton; a video in which MacMillan and McGrath explain the rationale and benefits of the approach, which they call "Discovery Driven Growth," at; their book; and McGrath's blog at Several readers have pointed out the parallels between their perspectives (developed over the last 10 years or so) and the "lean startup" notions of Eric Ries ( - these complementary perspectives are crucial to avoid the pitfalls you've identified. Another important and useful resource is what we call Agile Market Research. Entrepreneurs often make bold assertions re: market potential, e.g., how the market will respond to their product, conversion rates, projected ASPs, etc. These hypotheses and others can be tested in advance of significant investment and commitments. While "listening to customers" and "build it and (see if) they will come" can be informative, more accurate methods can be used to predict market response and are recommended when opportunity costs, investment and/or uncertainty are high. Data can be obtained and a predictive model generated relatively quickly and at lower cost, compared to a market test - the model is also more robust (e.g., allows for the testing of many different configurations, price points and business models, not just one or a few). While predicting consumers's response to "very new" products (e.g., iPad; Twitter; etc.) is fraught with challenges, it's not impossible (for further discussion, see Dr. Phil Hendrix, immr and GigaOm Pro analyst
Phil_hendrix is now following The Typepad Team
Jul 31, 2010