Why We Turn Down Some Projects
We decline work where no single workflow is economically material, where nobody owns the outcome, where the data can't be reached, or where the request is general AI curiosity with no workflow attached.

Most vendor relationships are structured to make “yes” the default answer to almost any inbound enquiry — there is a commercial incentive to find a way to say yes, even when the fit is questionable. We have deliberately built the opposite incentive into how we operate, and it is worth explaining why.
01. What we say no to, specifically
- “Show us what AI can do”, with no workflow attached. Curiosity about AI in the abstract is not something a Diagnostic can measure — there is no baseline to establish and no acceptance criteria to write, because there is no specific recurring deliverable to map.
- Volume too low for the arithmetic to work. The basic economics of a workflow fix depend on there being enough recurring hours at stake to justify the cost of measuring and building. Below a certain volume, even a perfectly executed build will not recover its own cost, and proceeding anyway would be dishonest about the numbers.
- Nobody owns the outcome. A workflow needs a specific accountable sponsor who can decide and unblock — without that, there is no one able to sign off on acceptance criteria, approve the design, or make the calls a build requires along the way, and the engagement stalls regardless of how good the work is.
- Budget that cannot carry both a Diagnostic and a build. Committing to a Diagnostic without the resources to act on what it finds sets an expectation neither side can meet.
02. Why we would rather say this upfront than discover it later
A project that should not have started tends to fail slowly rather than quickly — it drifts, it stalls, and everyone involved spends months discovering the fit problem that could have been named in a single short call. Saying no at the fit-call stage costs us a potential engagement. Saying yes anyway and having it fail costs both sides considerably more, and damages the trust that makes the rest of our process credible.
03. What “worth a conversation” looks like instead
The inverse of the four criteria above: one recurring deliverable that is economically material, an accountable sponsor who can decide, data and systems we can actually be given access to, and a real business trigger behind the request — a deadline, a cost, a growth constraint. This is stated plainly on the homepage as the specific bar a fit call is checking against, and it is checked by the form before any call is booked.
04. What this means if you are not sure you qualify
The free fit call exists specifically to answer that question honestly, without cost to you either way. If the fit is not there, you will hear that directly, along with the reason — not a vague deferral or a soft no designed to keep the door open for its own sake.
The fastest way to find out if your workflow is a fit is to ask us directly — discuss one workflow →



