How long should your MVP take?
Throughout the whole process of building OHNO, Poe and I had in our heads these dates we wanted to hit. I wanted to keep the MVP really small. 8 weeks, then validate. We did a good job of slimming down the features to a small set, but as you get going, you start to finesse the details, and a question starts to arise, when should you launch? How polished should the product be?
People like to argue this point, but mostly because it depends on how you define the terms; MVP, Validation, or Product. The literature is helpful, Lean Startup, Inspired, the usual suspects, but I thought I’d paraphrase what I’ve discovered from my own journey.
I like to think an MVP should be something a random person could use, instead of a customer you’re close to. Random people give you a true sense if your idea will work. They don’t really care about you personally, so they don’t bullshit you. And all the math will look real. If you look at the analytics from your ‘friends’ the numbers will look much more positive than the real world will reinforce.
There are a few things people have said to me over the last few years in relation to this;
If you’re not embarrassed by your MVP, you’ve launched too late.
This has always bugged me. Because I think if you’re embarrassed your customers might be too. And I don’t want customers to be embarrassed. I want them to talk about it at parties. When was the last time you bought something and said to yourself, “Wow? I’m so embarrassed to use this product. Awesome right? You should use it too.” NPS 10. You don’t want to feel ashamed of your work.
It should be just enough, no more, no less, to validate your product.
Okay. This I kind of agree with. So then what? Most teams want more than validation. They want to get the ball rolling and turn their idea into a viable business. You can validate a lot of business ideas with paper prototypes. So I don’t think this is clear enough advice for most people.
Generally speaking, my observations are MVP’s take way too long. I suspect if its over 3 months something is a bit off the rails. If it’s a hardware company or your building a spaceship or something — okay, I get it. No judgment.
You can usually validate your assumptions in weeks, instead of months. The way we did it at OHNO was to build a fully clickable prototype, and we sat down with potential customers and just asked them to use it. Then, we asked them to pay for it if they wanted. Which put them to a decision over the table for the purchase.
I think too often teams are building huge MVP’s that take months, even sadly, years, before they put it in front of a customer to find out they were way off the mark.
Perhaps the best answer I have seen to when you should launch is this;
When you have a skateboard.
Teams often think of MVPs like cars with no doors, and I would say this is more common with engineers. “It’s kind of a car, but yeah, it’s a bit fucked. There are no doors. I know that's bad. But it rolls and has an engine and that’s fine for an MVP.” I think this is a bad idea. Designers tend to be the opposite, they focus on the world’s most expensive skateboard. Gold plated trucks. Wheels designed by Kanye West.
But a good enough skateboard will usually do. Aim for that.
You want an MVP that does fewer things, really well. Like a skateboard. I think the above illustration sums it up best. And I know it’s done the rounds before but I have often shown it to leaders at software companies and they have never seen it, it needs more airtime.
