The diff that tripled

My first product owner job was an internship at a consulting firm. Big client, big project, and a quote signed before I even arrived: specs and development, one number, day one.

The spec phase took a year. I joined somewhere in the middle, taking notes in meetings where features got added the way you pile food at a buffet. Nobody flinched. The price was signed, so the price had to be right.

Then development started, and three months in we were late in a way I couldn't explain. Not “a sprint slipped” late — structurally late, like the plan belonged to a different project. I wanted to understand how you miss by that much, so I dug up the very first version of the spec and compared it to the one we were building against. By hand.

We had tripled it. Three times the spec volume, three times the features. Same price. The development was already drifting 80% past the initial quote, and it was going to get worse.

I was the intern. I flagged it anyway.

What happened next is the part I kept. The firm changed how it sold: two phases. A quote for the specs. Then a quote for the development, based on the specs that actually existed. They accepted that a client might take the finished specs and walk before phase two — some did. The projects stopped blowing up.

The first project ended up with the lawyers.

Years later, my studio sells a fixed-price scoping sprint first, and prices the build only from what comes out of it. It's not a methodology. It's that diff.