You post reminds me on a post I wrote a few months ago about peripheral vison:
http://www.cph127.com/2007/03/17/peripheral-capability-%e2%80%93-the-way-towards-better-product-development/

…I think we are on the same page here?

The top people on both sides of the debate expressed lots of grumbles about having to think everything through twice, with some asking for the CEO to make an “executive decision” for one scenario or the other. Eventually he did — perhaps after being lobbied privately by some members of his team. But the people whose scenario was discarded were never happy with the CEO’s decision, and continued to voice their opposition.

Essentially, the two scenarios were a mass market offering (same service for everybody) versus a portfolio of segment-specific offerings. Similar in some sense to the (apparent) difference between the Obama strategy and the Penn-influenced Clinton one.

“I am leafing through the February 2008 edition of Harvard Business Review, and there my wondering eyes behold “The Experience Trap” by Kishore Sengupta, Tarek K. Abdel-Hamid, and Luk N. Van Wassenhove. The authors studied experienced managers put in charge of complex software development projects. Their subtitle described what they found, “As projects get more complicated, managers stop learning from their experience.” Instead, they write,

‘managers find it difficult to move beyond the mental models that they have developed from their experiences in relatively simple environments or that have been passed on to them by others. When complications are introduced, they either ignore them or try to apply simple rules of thumb that work only in noncomplex situations. What they don’t do is materially improve the quality of their mental models to take into account the realities of complex projects.’”

“Thus it is that people with experience in one environment so often fail so dramatically when confronted with complexity in another, changed environment. Thus, for example, while Lyndon Johnson knew a lot about Texas and George Bush may have known a little, the one was faced with Vietnam, the other with Iraq. The results were not, and have not been, pretty.”

So, we then tried business planning (the company was still in pre-launch mode) on the basis of two views of the future, not one. Let’s take each scenario, we said, and run it right through to the end, keeping both scenarios alive as far as we can. In other words, since we can’t agree on the future, let us not try, but let us keep different views in play. This was surprisingly easy, technically. Most of the technical, marketing and financial plans could be made robust to the two different scenarios, with lots of options kept in play right up to service launch, and some even beyond them. This approach would have been very robust.

However, what caused trouble was not a technical or business planning issue, but a psychological one. Top management could not keep two views in play. Most of the people, some of whom were very bright and experienced, could not plan a business on two contradictory views of the future simultaneously. They could not do this even though the business was in planning mode, and would not be in operations for some years. Lower-level people had less of a problem with this, perhaps because they are used to working on things they don’t themselves believe or because they had less at stake in the outcome.

So the big implementation challenges with the ideas suggested here about asumption-finding and assumption-challenging are, I think, more likely to be human-related rather than technical or intellectual.

Cheers!
Elizabeth

Looking at this problem from a different angle, what happens when the blind side hit comes not from a competitor, but from the consumer? How do you forsee and adapt to consumers appropriating your marketing slogans or using your products in completely unforseen ways, both positive and negative?

None of this is to say you shouldn’t be on the look out, but I think today’s global marketplace and consumer empowerment calls for an adaptable, open-source model.

The problem with making explicit all our assumptions is that the list is infinite. That’s one of the big problems with top-down artificial intelligence–if you try to write down all the stuff that you, as a human being, assume when acting as an intelligent problem solver, you have a huge task. Everything from “big pills are harder to swallow” to “sick people want to get better” might be implicated at a pharmaceutical company, for example. Doug Lenat’s Cyc project is the closest thing to a brute-force approach to writing down commonsense knowledge, and they’ve been going at it since the mid-1980s without cracking the problem.