MVP stands for minimum viable product, and both adjectives are misread. Minimum does not mean poor, and viable does not mean finished. It means the smallest thing a real customer can use to do one real job, so that you learn whether the job is worth doing.
The purpose of a first version is to learn. Everything about its scope follows from that.
01Start from the riskiest belief
Every new product rests on beliefs: that people have this problem, that they will pay to solve it, that they will switch from what they use now. Find the belief that would sink the product if it were false, and build the least that would test it.
02One job, done properly
- Pick one kind of user and one job they need done.
- Make that one path complete: start to finish, with nothing missing.
- Leave out the second kind of user, the settings and the reports.
- Do by hand, behind the scenes, whatever is expensive to automate and rare.
A narrow product that works beats a wide one that half works. People forgive missing features. They do not forgive a product that fails at the one thing it claims to do.
03What must not be cut
- Security of accounts and data
- Working on a phone
- Speed: a slow product reads as a broken one
- A way for users to reach you
- Measurement, so you can see what people do
04What to measure after launch
- Did people finish the one job?
- Did they come back and do it again?
- Where did those who gave up stop?
- What did they ask for, in their own words?
- Would any of them pay, or pay more?
05Deciding what comes next
After a few weeks of real use you will know more than any amount of planning could tell you. Extend what people used. Fix where they stopped. Drop what nobody touched, however much you liked it. The second version should be shaped by evidence, not by the original wish list.
In short
- A first version exists to test the belief the product depends on most.
- One user, one job, complete from start to finish.
- Security, mobile, speed, support and measurement are never cut.
- Let what people actually do decide the second version.
Questions
How small is too small?
Too small is when a user cannot complete the job without you standing beside them. If they can finish alone, it is big enough.
Should a first version look polished?
It should look trustworthy. Clean, consistent and quick is enough. Decoration can wait; credibility cannot.
Is the first version thrown away later?
Not if it is built on sound foundations. The scope is small; the engineering should still be something you can build on.
Can Quantum Beetle build a first version?
Yes. We agree the one job it must do, give a clear quote before work starts, and build it so the next version extends it instead of replacing it.
Sounds like your problem?
Tell us about it. We'll say honestly whether the swarm can help, and what it would take.
Read next
- Website, web app, mobile app or custom software: how to choose what to buildSoftware development
- What a web application security test covers, and what to ask forSecurity testing
- How to get your company cited by ChatGPT, Perplexity and Google AI OverviewsAI marketing · GEO
- What a website costs: the eight things that decide the priceWebsites
- How to brief a software development company so you get what you meantCustom software
- Custom software or an off-the-shelf product: which one fits your businessCustom software