Killing What's "Kinda Working"

Probably the hardest decision of any business is to kill what’s “kinda working.“ I’ve lived through this probably more times than I can even count. At Coinbase we did this, many times. Decentralized exchange that’s kinda working… kill it. Margin program that did 500m in volume but regulators hate it… kill it. At my own companies, API that’s kinda working…. kill it. AI interviewer that’s kinda working… kill it. Data warehouse that’s kinda working, kill it… - By myself, I really haven’t gotten to anything that does definitely work, but I will.

There’s this saying, if it’s not a hell yes, it’s a no. That is true for anything business. It’s amazing how common it is, especially early stage, to kill business lines.

The flipside of this is Coinbase’s retail business: invest in this at any cost. Stablecoins that are growing 200% year over year: invest at any cost. Make. Those. Things. Work.

As engineers “kill it“ becomes almost a normal part of everyday life, especially in startups. Your life is a series of over and over again venture bets. You just need to ride the wave, unless you’re at a big company.

The really really hard ones are where it comes to “it’s kinda working, but double down anyway“. The problem is you invest so much up front it’s hard to give up. I saw this with businesses like Coinbase wallet, where it wasn’t obvious to invest. Who wants a wallet that makes no money and provides a little bit of value? These’s are product lines where it wasn’t a “hell no“, but we believed in it enough anyway to make it work. The founders were the ones who believed in this. Otherwise, it wouldn’t survive the corporate ladder. It’s better when there is no corporate ladder though anyway.

It’s hard to get there, and founders need to *believe* in something that “could work“. I believe that flyflow “could work“. It’s a weird place of believing my company “could work“, but also having light enough conviction to move forward when it does or doesn’t. This is all part of the game of pre PMF product.

Really the people who win are the people who have a high hit rate. A high hit rate of products that make it. It takes intuition. Intuition of the things that are “close to working, but not quite, but could“. It’s kind of a special zone where you don’t have “something“, but you don’t have “nothing“ either. These are the special zones that make for great products.