I usually start with a consumer problem, often one that I have experienced.  For example, how do I invest my money and beat other investment alternatives?  How can we improve the “user experience” of airlines (a thought I have every time I fly)?  How can we improve our government?  How can I collect data from other sources to harden my assumptions in a model?  I try to come up with at least one problem each day, which isn’t hard since I question everything.  I then use one or more of the methods below to solve the problem/s I have identified.  

1.  The world view.

This approach is based on some deeply held belief that you have about some under-appreciated truth.  For example, I believe that problems involving common knowledge (Karl Popper’s term) are ideally suited for distributed solutions (e.g., Wikipedia, markets) while those involving scientific knowledge (Popper, again) are best solved by extremely small groups of brilliant people.  

Steven Landsburg shares his world view in The Armchair Economist.

The world abounds with inefficiency, and to the untrained eye much of it seems to be the result of “cutthorat competition” or “markets run amok.”  But the Invisible Hand Theorem tells us that if we seek the source of inefficiency, we should look for markets that are missing, not for markets that exist.

We already do this implicitly when we think about problems.  I’m suggesting here that you explicitly develop a point of view and use it to solve problems you have identified.

2.  The logical extreme.

I typically take an economic or technical trend to an extreme and ask, “what if.”  For example, storage costs are decreasing faster than Moore’s Law.  What if storage is free?  If you could provide infinite storage at no cost, how would that change things?

What if GPS chips are embedded in every phone, computer, and car in the future?  What if those devices could pinpoint the location of every consumer on the planet at anytime?

China has moved 250MM people into urban areas over the past few decades.  What if they move another 300MM into urban areas?  How does that change things?

3.  Less is more.

This approach has been popularized by 37 Signals and Evan Williams.  College students are using social networks.  What if you just provided a social network for them?  Blogging is crazy popular, what if you had a 140 character blog?  What could you do?  

The idea here is to look at bloated products and pick off one small user segment or feature and do it really, really well.

Next up, The Team, Part I:  Who do you need, and when do you need them?