Why you should not be building a minimum first version

Tech startups have inherent iterative potential: you can be one thing one day (say a Group payments company), and something completely different the next (for example an alternative Payments network). A school of thought has sprung up around this called Lean Startup Methodology. One of the key principles is that you should focus on building a Minimum Viable Product. The main reason is so that you don’t spend too long building out the first version before you realise you should be building something completely different.

This has become pretty widely accepted and almost every early-stage startup I speak to tells me that they’re focused on building their MVP. However, a trend that’s started to emerge is that what many of these guys think is their MVP is actually a Minimum First Version. On first thought it might seem like a piece of sophistry to try and distinguish between the two but I think there’s an important nuance.

As soon as you start thinking about Minimum First Version you start thinking about your Final Version, and therefore what a stripped down version of that might look like. That is often radically different to your MVP.

To give an example we started out in the Group Payments space. We wanted to build a tool that would make it easy for people to collect money from lots of people. The end product at its core would be a site where they could set-up groups, issue bills and reminders, and see a summary of information. We couldn’t handle payments ourselves yet, so we thought our MVP/Minimum First Version (we treated them synonymously) would be an ugly version of that which used the PayPal API, and had only this core functionality. To be fair, we built it in three weeks. But because it was still quite a substantial product we then had to iterate and improve the product over a number of months to get everything to work as it was supposed to, to get a fair reflection of what users thought of it.

We learnt that there was a kernel of the product people found really useful, and that larger groups and businesses found it really useful, but did we really have to spend 4 or 5 months finding that out?

No. Last week we realized what our MVP was and built it in a day: a bit.ly type link generator for payments: you say how much you want and what for, and we generate a link. I then thought back to the original list of people we were trying to help and it would have solved the core problem for all of them.

That was our Minimum Viable Product. We should have built it in a couple of days, shipped it, and then focused all our resources on getting people to use it and seeing what worked and what didn’t. Instead we had a much more complicated product, which needed months of work to get the various bits to work, before we could have those same conversations properly, and learn exactly the same thing.

Now when I think of the number of startups I know who have spent upwards of a year just to build their first product, I really worry if they’ll ever even find out what that kernel is for them, and even if they do, if they’ll have the time to iterate on that sufficiently to get to product-market fit.

So here’s my question to you: Are you building a Minimum Viable Product or a Minimum First Version? Or, more concretely, what’s the equivalent of our link generator for your business?