Central Perk: 8 apr. 2011

vineri, 8 aprilie 2011

Seth's Blog : Insist on the coin flip

Insist on the coin flip

Very often, we're challenged to make decisions with too little information. Sometimes, there's no information--merely noise. The question is: how will you decide?

Consider the challenge we faced when setting the pricing for a brand of software we were launching in 1986. It was the biggest project to date of my short career... more than a year of work by forty people. Should these games cost $29, $34 or $39 each? My bosses and I had one day to finalize our decision for the salesforce.

Unlike Harvard case studies, we had no graphs, no history, no data. We were the first in the category and there was just nothing significant to go on. The meeting was held late on a Thursday. In addition to my newly minted Stanford MBA, we also had two from Harvard, one from Tuck and another I think from Wharton in the room.

We talked for an hour and then did the only intelligent thing--we flipped a coin. To be sure we had it right, we double checked and flipped two out of three. The only mistake we made was wasting an hour pontificating and arguing before we flipped.

This is also the way we should settle closely contested elections. We know the error rate for counting ballots is some percentage--say it's .01%. Whenever the margin is less than the error rate, we should flip. Not waste months and millions in court, we should insist on the flip. Anything else is a waste of time and money.

Or consider the dilemma of the lucky high school student with five colleges to choose from. UVM or Oberlin or Bowdoin or Wesleyan or who knows what famous schools. Once you've narrowed it down and all you're left with is a hunch, once there are no data points to give you a rational way to pick, stop worrying. Stop analyzing. Don't waste $4,000 and a month of anxiety visiting the schools again. The data you'll collect (one lucky meeting, one good day of weather) is just not relevant to making an intelligent decision. Any non-fact based research is designed to help you feel better about your decision, not to help you make a more effective decision.

One last example: if you know from experience that checking job references in your industry gives you basically random results (some people exaggerate, some lie out of spite), then why are you checking?

When there isn't enough data, when there can't be enough data, insist on the flip.

By refusing to lie to yourself, by not telling yourself a fable to make the decision easier, you'll understand quite clearly when you're winging it.

Once you embrace this idea, it's a lot easier not to second guess your decisions--and if you're applying to college, you'll free up enough time to write a novel before you even matriculate.