Hacker News Reader

Comment by ntoshev | Hacker News Reader

PG says the first iteration doesn't matter much as long as the founders are good and iterate quickly. Empirically, is this the case? Airbnb and Dropbox certainly didn't iterate on their concepts that much. Google's first shot was excellent. Facebook captured 90%+ from Harvard students before moving on to conquer the world. Are there real counterexamples, or even better, hard data?

(If only YC recorded and aggregated key metrics from their startups, it would be so much easier to answer such questions)

View on HN

Airbnb was iterating prior to being part of YC, they've given several interviews where they talked about how they'd gone through several launches and failed to take off and had to iterate on their model. For example they started out by only being an introduction service, but they found handling the payment in person was actually a major issue (socially awkward, no record of transaction, etc.) so the end up incorporating payment handling into their service. That was likely a critical iteration.

Google's primary business is Ads which wasn't what they started with.

Handling all payments is absolutely critical because it leads to a review/rating system that is credible (which is critical for a service like Airbnb where strangers are offering and renting space to/from each other). As soon as the payment processing is gone you end with a review/rating system like Yelp. Namely one this is easy to game (for example business owners offering free things to people who give them a good Yelp review) and therefore carries little to no weight. I've used Airbnb three times now. The credible rating system was crucial for me to feel comfortable staying with a total stranger.

Airbnb's first iteration was presented as being much more about actually renting air mattresses and couches in people's home. They iterated into ebay for spaces.

Dropbox's first iteration was a one-guy team working on the problem. The founder iterated into a two-founder team based on feedback. I'd say that's a significant iteration.

If you read The Facebook Effect, Facebook was simply the latest in a line of "fun hacks" Zuckerberg unleashed on his fellow students at Harvard, each one building on the next.

Sounds like every one of these projects involved iteration. It's fun to think awesomeness springs fully-formed from one's forehead, Zeus-like, but I don't that happens to often in practice.

Replace "iterate" with "kept working on and improving" and your comment makes just as much sense: any company, old or new, evolves and tries out new things and abandons other things that don't work — in that sense, saying iteration is crucial to startups is pretty much just a tautology.

I think PG was hinting at the fact that for startups, iterating often means pivoting, so your first iteration doesn't necessarily lock you into path dependency. Which is a lesson that doesn't really apply to Facebook or Dropbox or Google, because they got the general direction right the first time.

Data would be helpful, I agree.

But sometimes in iterations, you end up totally changing the product. So you might be thinking this is the first iteration of Groupon but it wasn't, the first iteration was The Point which really didn't work out and no one knows about it.

Prior to September 2006, Facebook bore little resemblance to today's newsfeed-centric iteration.

PG had a line recently like -- "we're looking for teams building BASIC for the Altair" -- an obvious reference to Microsoft. And Gates and Allen had a business before that analyzing traffic data or something.

Reply on news.ycombinator.com