An interview with Paul Graham on TechCrunch is attracting flak from the blogosphere for the investment criteria he uses for Y Combinator.

Last week we witnessed the sale of Kiko, a well-designed Y Combinator-funded online calendar. The deal went through via eBay to an as-yet-undisclosed party, for a quarter of a million dollars. Which is a good return for Paul and the Kiko founders, but not the really big bucks that they would have seen had Google bought it. Google of course launched its own calendar app, integrated into Gmail, which prompted the Kiko sale.

So does a fire sale on eBay, albeit a profitable one, mean that Graham is any more correct when he advises startup founders not to worry about business models?

I have previously criticised Graham for making noises about the lack of importance of a business model when starting up, which is pretty much the opposite to the advice we offer to internet entrepreneurs. Yet we all agree on various things: we’re not in a bubble, disruptive technologies rock our worlds, and accepting VC money too early can lead to (years of) Pain.

I guess the immediate caveat is that if this strategy works for Graham, it works for Graham. But because he is viewed as a thought-leader (he’s a published author and renowned speaker) he is listened to, so there is a very real threat that some entrepreneurs will accept his opinions as sound facts. Which works to some degree if you have a Y Combinator investment, but not if you’re bootstrapped and investing your time into a project that needs to pay off, somewhere down the line.

In the TechCrunch interview he says: “What I tell founders is not to sweat the business model too much at first. The most important task at first is to build something people want. If you don’t do that, it won’t matter how clever your business model is.”

Sure, but pinning on a business model retrospectively isn’t going to necessarily help either. It may be that there isn’t much of a business model, apart from advertising. Adding a bit of Adsense tends to be the lowest common denominator, and Adsense only works well in niches (the best Adsense websites plan on attracting a niche audience…).

I think you need to define your target audience / user base, then roll out apps / websites that encourage them to visit and interact. Build scale while collecting the right sort of data, for use later. That means not necessarily opening up all of your features for unregistered users (this is probably an unacceptable statement for the 2.0 Brain Police but how else are you going to profile your audience?).

“I get a lot of criticism for telling founders to focus first on making something great, instead of worrying about how to make money. And yet that is exactly what Google did. And Apple, for that matter. You’d think examples like that would be enough to convince people.”

Startups that become Google or Apple happen extremely rarely, so this doesn’t entirely stand up. It took Google almost 5 years before it executed Adwords, which now accounts for 95% of its revenues. Without that brainwave it could have been vastly different, right?

Still, he’s got his heart in the right place:

“One startup I dream of funding is the one that kills the record companies. You know your business model is broken when you’re suing your customers.”

Hear hear. Let’s hope this is what the Kiko founders are working on!