How we choose what not to take on

We guarantee the ship date or the money comes back. A guarantee like that is not held up by optimism. It is held up by everything we say no to. Here is the filter, in the order we apply it.

Can we scope it on day 0

If we cannot write a fixed scope and a fixed date by the end of the first day, it is not a sprint. Some work genuinely cannot be scoped until you are inside it. That work is real. It is just not fixed-price work, and pretending otherwise is how you end up issuing refunds.

Is the surface area bounded

A good sprint touches a knowable set of files and systems. "Add metered billing" is bounded. "Make the app faster" is not, it is the whole app. We will reshape an unbounded ask into a bounded one with you, or we will decline it.

Does it need people we cannot put in a room

Some projects need on-site stakeholder workshops, a week of discovery with your sales team, a designer in the room. That work is valuable. It is not what we do, and a 14-day remote sprint cannot fake it.

Every sprint we decline on day zero is a refund we do not issue on day fifteen.

What this costs us, and why it is worth it

Saying no costs us real revenue every month. Sprints we could have signed, walked away from. But the refund rate is zero, the average ship time is 11 days, and founders send other founders. That trade is not close. The discipline of no is the product.