I would say those two are ongoing performance metrics to track, and not PMF criteria. If you were to buy a car for example, safety/uptime are table-stakes.
One question: You can achieve PMF via rapid build of a prototype or MVP, test it with users and learn on it. Isn’t SpaceX doing the same just in virtual environments and maybe with less external testers? With all the simulations, isn‘t it very close to the software based build-measure-learn cycle?
It is pretty close, but given that hardware is such a key part of flying rockets and that hardware aspect cannot be run through build-measure-learn cycle as fast, it still ends up being very different from pure software iteration cycles.
additionally success criteria for launch service could be :
1) Safety ?
2) minimum uptime : in case of any on flight issue time taken to fix it?
I would say those two are ongoing performance metrics to track, and not PMF criteria. If you were to buy a car for example, safety/uptime are table-stakes.
Thanks for this, very interesting article!
One question: You can achieve PMF via rapid build of a prototype or MVP, test it with users and learn on it. Isn’t SpaceX doing the same just in virtual environments and maybe with less external testers? With all the simulations, isn‘t it very close to the software based build-measure-learn cycle?
It is pretty close, but given that hardware is such a key part of flying rockets and that hardware aspect cannot be run through build-measure-learn cycle as fast, it still ends up being very different from pure software iteration cycles.