Sometimes the polish is all that counts. You can’t judge the market for a five star hotel by building a seedy motel and seeing how well it performs. In some cases the quality of product is more important than the type. When the interface is the killer feature, it’s tricky to go “lean”. If your belief is that people will appreciate a fully polished beautiful to-do list, you can’t show them a scrappy UI to test the market.

On creating good internal tools that make it easier for you to validate your product experimentation ideas.

Best practices sound good in isolation, but they can suck the life out of developers.

Nate Bolt of Bolt|Peters invited me to speak at User Research Friday last week. My obvious first reaction was "user research? We don't do any." Nate was interested in the perspective of a design firm that has no formal process for user testing or research, so I did my best to coalesce a bit of our most recent thoughts on astonishment and novelty for the packed house at their SOMA offices.

Effectively, they're a watered-down version of the Google "20% time" scheme: in every sprint, everyone at FP gets to take a day working on something of their own devising; with 2-week sprints, this works out as 10% of our time spent on R&D. We factor the time they take into our velocity calculations, so these days are included in our planning and accounted for.