Gary Flake recently announced the founding of MSFT Live Labs. I found his manifesto (via Richard McManus) for Live Labs to be an interesting departure to most “labs” or “skunkworx” groups within large corporatopns.
Most labs takes the “northeast” approach to innovation... riskier, more disruptive innovation, and longer time horizon than typical product develpoment/management innovation process. As a result, the lab/skunkwork unit is usually as far away from core products & business as possible (without losing context/vision) in order to be able to innovate independently without the baggage of corporate scared cows. Distance from the mothership, in many ways, is hugely desired for both political and innovation sake.
Gary, however, is using a more strategic approach to innovation... 1) identify stakeholders, technology, strategies, and products, 2) Map all variables within a vector space (geek speak) 3) identify “whitespaces” 4) create solutions to fill in whitespace 5) add value by creating synergies.
For anyone that have taken an MBA course on innovation and read Innovator’s Dilemma, this is hugely sacriligious. For those that adhere to the Michael Porter’s idea of “strategy = trade-offs on the efficient frontier” this is also very disconcerting. Academics do not believe someone should be able to pull this off. To be able to innovate significantly while PARTNERING with existing legacy business units is close to impossible. Many have tried and gotten dragged back to the daily grind of politics, meeting earning estimates, and managing cannibalization.
BUT if you really think about it... this is the ONLY way to innovate relevantly. Too many innovation groups end up creating solutions to non-existant problems (nano blah?) OR solutions mis-understood by the mothership (Xerox’s anything in the 70’s). This is a very brave approach to innovation... to ensure relevancy while leaping toward step function changes... I wonder if this can work... I’m sure there are a few projects that does fit the definition... better ad serving algorithms, better presonalization, recommendation... anything that focuses on optimization of an existing business or product function could all possibly work... the key question than is: in the long run will these advantages remain “strategic” and sustainable... or algorithmic superiority will go the way of TQM - price of entry but not “strategic” as defined by Porter and eventually get competed away...