Monday, May 26, 2014


When one person is starting out with an idea, the primary risk is he'd drown himself into the implementation of it and forgot to reflect on how sensible it is, and whether he'd chosen the most effective way to implement it.

When a group of people are starting out with an idea, the primary risk is they'd become unable to arrive at a consensus. Even if one of the group members had premonition ability and predicts accurately what'd happen in the next 5 years - he'd have a hard time to convince his cofounders that things would happen as he predicted.

Yet in a group setting, having consensus is always needed to make effective decisions. Otherwise you'd be just wasting your time delaying the day of reckoning - the day you realize you don't really have a coherent group, and thus it's meaningless to work together.

Thus, at the idea level, for a startup - there're two major risks:
  1. You have perfect consensus (e.g. only 1 founder), but you don't verify if that's the right thing to do
  2. You have no consensus (e.g. 3 cofounders, each believing in different things), and instead of facing this problem head-on - you delay the day of reckoning to keep the peace
Once explained to this point, the derivation of next steps should look so trivial to look almost like I'm speaking in tautologies. Before an idea has taken off in the market (e.g. making yet another convenience camera app before Instagram took off), if I explain it to you - it should normally sound like a bad idea. You would tell me it's a bad idea. And if I'm inexperienced, I might be led to believe that it's a bad idea. When you tell any investors about it, they're almost guaranteed to tell you it's a bad idea. And now you'd feel really right that anything like Instagram is a fucking bad idea - except that it makes a billion dollars a few years later.

And that's already ignoring the needed discussions about how to implement the thing. I'm already assuming perfect execution afterwards. Idea, is just one of the moving parts. More moving parts, more chance of failure.

So then, the reasonable expectation for a startup, is to start from a bad idea which you still have some suspicion to work - and then produce a logical and consensual process to refine it into a good idea via market response. Before you have a correct answer, the focus should be to arrive at consensus that you don't have the right answer yet, but that you have a good process to find out.

As I said... this should be so trivial among members of startup circles now the derivation should sound like a tautology. It's just the party line of any startup conference, basically. So why am I still typing this?

Because a friend of mine asked about this something like a month earlier. Everyone in the Valley should know the rule above, yet knowing the rule (from books, talks, advice from advisers, etc.) doesn't mean you know how it applies to reality.

"If I tell you of my ideas right now, you won't really believe in them."
"I think I believe in you, Martin. You're more experienced."

Well, no. First, you won't believe me if I laid it out to you. Second, I already know my predictions won't be perfect. Third, convincing you is not the point. The ideal case is you help me find the flaws in my ideas and we work out how to arrive at the correct answer.

But that, really, is still not the point.

You're a human, the really important thing for you to know - is how you'd feel while you're trying to do that.

You'll feel uncertain.

And you simply need to accept it, and be brave.

(But always hedge your risks - that's why you can be brave in the first place :P )