3 ms·

An MVP should be built in around one week. 90 days is definitely too long, and, if it takes you that long, chances are that you're getting too invested. I would

by dvt 7y ago

An MVP should be built in around one week. 90 days is definitely too long, and, if it takes you that long, chances are that you're getting too invested. I would say you should dedicate at most 30 days to build an MVP. Longer than that and you're probably adding bells and whistles that aren't relevant to your core value prop. But I agree that it's tough and you'll often second guess yourself: "Will people really pay 10 bucks a month just for a better S3 interface?" -- Dropbox says yes. Last year, my goal was to build something every two weeks[1]. It was really fun, but also really exhausting, and, for more complicated projects, two weeks was definitely too little time. But it's a fun experiment I'd suggest everyone should try. [1] https://dvt.name/2019/01/06/retrospective-stuff-2018/ https://dvt.name/2019/01/06/retrospective-stuff-2018/

walesj 7y ago
That's a pretty cool experiment. Thanks for sharing. I'm gonna check it out.
nostrademons 7y ago
The interesting thing about DropBox was that their "MVP" was a video, and then actually building the first viable product took 4 engineers 1.5 years. I would say that there aren't hard & fast rules to any of this. It took Google 2 years to build their MVP, and they also had multiple grad students working on it. The only "rule" to startups is that you have to survive long enough to get to the point where lots of people want what you've built. But there is a significant advantage to speed. Letting an MVP drag on for more than a couple weeks introduces a lot of different failure modes. You get demoralized and give up. You get too attached and fail to pivot when the initial version doesn't work. You build something too complex for initial users to understand. Technology changes and makes all your past decisions obsolete. The market changes and the problem you solved is no longer relevant. You run out of money.
- dvt 7y ago
  This is a good point. To wit, I'd say the Google MVP was probably the Stanford PageRank paper.

by dvt 7y ago

walesj 7y ago

That's a pretty cool experiment. Thanks for sharing. I'm gonna check it out.

nostrademons 7y ago

The interesting thing about DropBox was that their "MVP" was a video, and then actually building the first viable product took 4 engineers 1.5 years. I would say that there aren't hard & fast rules to any of this. It took Google 2 years to build their MVP, and they also had multiple grad students working on it. The only "rule" to startups is that you have to survive long enough to get to the point where lots of people want what you've built. But there is a significant advantage to speed. Letting an MVP drag on for more than a couple weeks introduces a lot of different failure modes. You get demoralized and give up. You get too attached and fail to pivot when the initial version doesn't work. You build something too complex for initial users to understand. Technology changes and makes all your past decisions obsolete. The market changes and the problem you solved is no longer relevant. You run out of money.

dvt 7y ago
This is a good point. To wit, I'd say the Google MVP was probably the Stanford PageRank paper.