Excellent, clear thinking; thank you for this analysis. I wanted to ask more about: we both totally failed at... actually helping people. Changing people's b

Excellent, clear thinking; thank you for this analysis. I wanted to ask more about: we both totally failed at... actually helping people. Changing people's behavior is really hard. Yes, but it seemed to me -- and this post confirms -- that Wesabe was genuinely thinking about this problem of how to get people to change behavior, to improve their lives. Mint seemed (and still seems) to have a much lower ambition in this respect. Question #1: Was Wesabe actively engaging in experiments in how to help people/change behavior? I know the support community did help some people, but it sounds from your post that in its existing form it didn't do something palpably different from what people would do on their own. So it seems that Wesabe ought to have been building as many experiments as possible, hoping for something that seemed (at least anecdotally, at first) to help some subset of users above-and-beyond what they were able to do elsewhere. I suspect Wesabe was running some experiments but too few. If that's true, what prevented those experiments from happening? Infrastructure work/scaling challenges/security concerns? Was the employee or user count too high to efficiently run experiments? Question #2: If Wesabe was running a good number of experiments, why was Wesabe unable to secure investment for a longer runway of more experiments? After fixing the Yodlee problem (which I agree was a major problem), Wesabe seemed well positioned to lead in experiments to change behavior. The space still seems enormously ripe for innovation. So why did new funding not happen? Was Mint's success demotivating? Was the timing bad, given the economic downturn?