Eight MVPs in, I stopped calling them MVPs

By the time I'd shipped eight of them at Accenture, I'd developed an uncomfortable pattern-match. Some MVPs worked. Some MVPs 'worked' — in the sense that they demoed well, hit the acceptance criteria, passed user testing — and then died in production three months later. The code was identical quality. The users were equally real. What was different wasn't the product.

What separated the ones that lived

Every MVP that survived had a specific person on the customer side who had their reputation attached to the rollout. Not the sponsor who signed the statement of work — that was always someone senior. The operator two levels down. The one who'd be running the morning stand-up with the field crew on day one.

The ones that died had a sponsor, an acceptance process, a steering committee, and no operator. The product shipped cleanly into an operational void. Three months later the spreadsheet workaround was still in use because no one had stayed to make the new tool stick.

• Lived: someone specific whose KPI depended on adoption
• Died: adoption was everyone's responsibility, therefore no one's
• Lived: weekly ops review included the MVP for the first 90 days
• Died: the project team moved on once the signoff was complete

Why this reframes how I think about 'done'

A lot of shipping-focused advice stops at shipping. Ship the MVP, get it into production, move on. The problem is that the shipping act is the cheap part. The expensive part is the two months that come next — when the tool is live but the old workflow is still running in parallel and everyone's quietly still using the spreadsheet.

The MVP is the excuse to do the adoption work. The MVP isn't the product.

For founders

When a founder tells me 'we shipped the MVP but it's not getting traction', the question I ask first is not about the product. It's 'who on your customer side has their name on the rollout?' If the answer is 'our champion loves it' — that's a sponsor, not an operator. Find the operator. If there isn't one, the MVP isn't the problem.

Shipping a product is a technical event. Getting it used is an organisational one. They need different people and usually take different amounts of time.