We take on a small number of engagements at a time, on purpose. Every system we ship comes with a standing commitment, we stay involved while it runs, and that commitment only means something if the list of systems stays short enough for us to honour it.

So we decline more briefs than we accept, and we try to do it usefully: quickly, in writing, and with a referral when we know someone better suited. The most common reasons are fit, not quality, the problem is really a staffing problem wearing an automation costume, or the timeline requires skipping the diagnosis that makes the work safe, or the client wants a tool handed over rather than an operation improved.

Every yes is a promise to stay. The queue stays short so the promises stay true.

We publish these constraints instead of negotiating them privately because it saves everyone time. If you read this and recognise your brief in the reasons above, we are probably not your agency, and if you read it and think “good, that’s what we want from a partner,” the conversation usually goes well.

FILED 2025.07.07 · GRUNTFLOW RESEARCH