In 1st year of airbnb they had 2 bookings. 2! Only 2. Wow! But that would change once they understood one thing: you should do things that don’t scale to get your first customers. This is just one example of things they did that didn’t scale.

During Y Combinator the airbnb founders started to talk to their users (mainly in NYC). They learned many things about perfecting the user experience, but one of the biggest insights was: people have amazing places to rent, but their pictures suck. They don’t have great cameras and certainly don’t know how to take great pictures.

So the founders asked, “What if our hosts could just push a button and a professional photographer would just show up and take pictures for them?” That would be the perfect experience right?

So normally when you discover a feature need you immediately ask, “What will it take for development to build this? How much? How long?” And then you wait. Of course, other features may have priority so you have to put this idea in a list and who knows when it’ll get built.

But, if you have the internal mindset to do things that don’t scale you might just ask yourself, “How can we fake this and test it? How can we get it out in the wild immediately?” This is what airbnb asked.

So they did everything manually. Here’s how it broke down: 
1. ask current hosts via email if they’d like this complementary service 
2. make a spreadsheet with all the info 
3. send yourself as the “professional” photographer 
4. manually upload the pictures for the host

Boom, this feature is now “built.” 

But what happens when things start to get out of control? You can only do so much manually, right? Wrong. You delegate. You hire a part-time person who manages the process. Then as it gets bigger, you hire a full-time individual to manage other part-time or as contractors. This is what airbnb did. And then ONLY when things become too hairy and complex you build the feature (or maybe before you hire a full-time person).

Takeaway: Fake it before you build it.