How to filter startup ideas
I have a spreadsheet with hundreds of startup ideas. I've also tried to start several companies that were premised on ideas that, in retrospect, were bad. Now I have a simple framework – two tests that an idea needs to be pass before I consider it more seriously.
The honeymoon test. Few ideas are exciting after the first 2 weeks. In that initial period you're in a honeymoon phase – the positive parts of an idea consume you and you leave no space for doubt. As you dedicate more time to research, you talk to potential customers, research competitors, and figure out what execution looks like. Most of the time, excitement fades away. The takeaway: give yourself at least 2 weeks before fully committing to an idea and going into build mode.
The IPO test. Imagine a world where your idea achieves its full potential. Would you be proud of the impact you've had on the world? I made the mistake once of pivoting to an idea that I didn't really love.¹ If you're going to embark on a journey for 10+ years, you should pick something that you will be proud of. The startup journey has many highs and lows, and working on something you're proud of is the best way to keep your head up during the lows.
¹ The company was called Lightout. We installed digital screens on Uber cars to help drivers make money and help advertisers run hyperlocal geofenced campaigns with a self-serve advertising platform. I loved the first part of the idea but hated the second. At our peak, we had about 250 cars in the Bay Area, and it really was exhilarating to see our product in the wild. But deep down, a part of me always felt that the idea taken to its extreme – cities filled with digital billboards – was not one I would be proud of.