My favourite part of this was 10-15 calls a day with potential buyers, and they kept changing what they were offering until responses went from “that sounds cool” to “when can I get a PoC?” presumably without a line of code being written.
Having launched a couple of dead startups that started with several months of writing code first, this way definitely sounds better.
> Having launched a couple of dead startups that started with several months of writing code first, this way definitely sounds better.
This is basically what startup 101 tells you. This is what every successful entrepreneur will tell you. This is what every coach tells you. This is what every entrepreneurial book or blog will tell you.
But, this is also what every tech entrepreneur will ignore anyway.
This is one of those things that you have to experience a few times before you look back and think 'oh... they were right'. But coding is comfortable and cold calling is very scary. It's also against our nature to ask anyone what they think of your idea, because it might shatter your dream.
YC Startup school nailed this in one of their talks, the presenter opened the talk with something like "this is important advice that you will all ignore, and that's okay, my goal is to make you recognise the situation after you'll inevitably make one of these mistakes".
I'm not being a snob here. Trust me, I made this very same mistake. I ignored all the advice and poured years into building products that nobody wanted.
100% agree, and of course, I absolutely knew this going in. Not building my next thing until I have something someone will start paying for before it’s even built
The mystery is what happened between that phone call and the $100M ARR. The customer says "Can I get a PoC" but you don't actually have any code yet. You just hope your tech team is able to conjure whatever you were able to sell?
Enterprise software rollouts can take months to actually get started from the point of procurement.
This happened at one startup where the sales team bid on a RFP, won, and then had to build it while finalizing the deal.
(First cut ended up being trash and crashed as soon as the customer took it global. It was replacing a paper process and had worked fine in a small scale pilot with one sub org. Customer ended up going back to paper and it took 4 years to rectify and try again)
Having launched a couple of dead startups that started with several months of writing code first, this way definitely sounds better.