Hacker News new | past | comments | ask | show | jobs | submit login

MVP: My demo ended up in production and now were trying to scale SQLight to 1000s of concurrent users help!

If you want to build an MVP im all for it. It is 100% throw away. Don't try to "add a feature" don't try to "expand" ... You cut corners in design and spec Im gonna cut them in engineering. Let's all take the lessons we learned and build something good after.




Sure, we can define MVP to mean prototype and then it makes it a prototype, but is that useful?

Minimum viable product still means it's a viable product. If it's 100% throwaway, it wasn't viable.


Fair:

MVP lauches and the PM is at my desk. Hey we need these 5 features that you told me I would want but I swore would never be in the product.

MVP launches and the PM is at my desk. Hey this thing that I would have found if I took a paper prototype out on the street and showed it to random people would have been apparent ... well now we need that.

Hell I have told PM's that their print on demand product needed a dirty words filter. They told me I was wrong. Guess what the first feature was post launch in a panic. It's not like I haven't said "If you're launching in Germany you need to take checks" and then had someone come back when the sales numbers were dismal till they started taking checks. (And dear god someone tell me that Germany is past this).

Candidly most MVP's miss the minimum and viable because the people who are defining those things are making guesses with out doing a lot of research. Ethnography, Usability, all these things are much cheaper than development, and yet we dont do them...

Fuck it were doin it live.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: