Why we decline work.
Turning down an engagement is a commercial decision, not a posture.
Atlas has declined more work than it has taken. That is not a filter applied for effect; it follows directly from how we are paid. An engagement tied to an outcome that never arrives is not a small loss, it is months of work with nothing at the end of it, so the decision about whether we can move the number is made before the engagement exists rather than during it.
Three things usually decide it. The first is whether there is a real product with real customers — we work with established companies, not with an idea looking for its first sale. The second is whether anyone on the other side can decide. A business where no individual can approve a change will consume a year in meetings and finish where it started. The third is whether the economics leave room: if the margin on what we would generate does not cover the work required to generate it, saying yes is a favour to nobody.
What we do instead of a proposal is read the business first. Send the website through the intake and the answer comes back either way — including the answer that we are not the right firm, which is a real outcome of that process and not a formality.