Here is the million dollar question (or with inflation the billion dollar question).

As you know, I spent almost 3 years creating Verifiable.com, a data visualization tool. For 3 years, we failed to effectively engage the users and ended up guessing what they needed or wanted. Sometimes by watching what they did or had trouble doing and sometimes by seeing what they didn’t do.  It was very very frustrating.  We never did “guess” what the users wanted.  Possibly because there was no “user” to understand, maybe all of the users were so different no one solution was going to make them all happy, or maybe guessing is something only Steve Jobs is good at.

My new business, SaneBox,  was created explicitly to avoid the guessing game.  It is meant to solve a problem that people I knew wanted and needed solved.  They were getting too many envelopes in their Inbox and it was too hard to distinguish important stuff from stuff that could wait.  And here’s the beautiful thing: several of them had blogged about what they wanted.

So I built that.  The thing they wanted and I invited the friends that were having the email issues to use it. And found that some of it worked for them and other aspects didn’t.  So we iterated.  And iterated.  And iterated.

Now I have a “bunch” of users.  And I am still committed to asking each individual what they want, what they like, what they hate.  So the group breaks down to about 33% that I haven’t been able to get engaged in a dialogue, 33% that will engage lightly and 33% that I feel like adopting as part of my family. The last group are the kinds of people that send a periodic message with suggestions, thoughts, encouragement.  They will take the time to snap shot an envelope that they think was mis-filed and drop me a note about it.  Just the stuff you need to push forward.

Sounds great you say. No more guessing. Sure.  Except we are still a tiny team and the ideas everyone has had (including us) are big ideas that require loads of typing and and, wait for it…,  lots of iterating.  Remember that Just because someone thinks they want it today doesn’t mean they will love it when they use it.  So, how should we decide what to do next. 

So, in an ever increasing attempt at transparency.  Here is my current decision calculus in order of importance:

1. Increase accuracy.  With each new user, we find new edge cases.  Treat each one as crucial.  If one user has the issue, others will.
2. Easier to use.  Where ever possible make the product more obvious and less intrusive.  If one user stumbles, get rid of the bump.
3. Faster.  I am committed that when we do start charging for this service it will be priced as reasonably as possible.  The only way to do this is to make servicing each user as light as possible.
4. Add features that users can use right now that seem to fit the paradigms of 1-3.
5. And, of course, all ideas must pass the VSCF test: Vision, Schedule, Cost, Features!

I’m gonna save the actual ideas that we are banging around for the next entry.  I thought it would help to lay out the meta issues first.