AI readiness assessment · from $2,500 · 1–2 weeks

An AI readiness assessment that ends with a number, not a slide deck.

Ours is called the AI Impact Diagnostic. In one to two weeks it answers three questions: which of your workflows is actually worth automating, what that workflow costs you today, and whether the return justifies building anything at all.

That last one matters more than it sounds. Most assessments are written by people who get paid either way. This one can conclude that you should not spend the money — and if that is the answer, you will have paid $2,500 to avoid spending $15,000 badly.

What we actually do

Three things, in this order. No workshops, no maturity matrix, no scoring you out of a hundred against companies that are nothing like yours.

1. Map the workflows

We look at how the work runs now — who touches it, what they use, where it waits. Not the process diagram somebody drew two years ago. The one that is actually happening, including the spreadsheet nobody mentions in meetings.

2. Pick the one metric worth moving

One. Not a dashboard of twelve. If a project cannot name the single number it exists to change, nobody can tell afterwards whether it worked — which is how most AI pilots end up quietly unmentioned rather than obviously failed.

3. Record where that number stands today

Cycle time, cost per task, error rate — whichever one the workflow turns on. Measured, written down, and signed off by you before anything is built. This is the part almost everybody skips, and skipping it is why so few AI projects can prove they returned anything.

What you have at the end

A baseline you have agreed to, in writing: the workflow, the metric, and its current value. That document is what makes a later claim of improvement checkable instead of a matter of opinion.

You also have a straight answer about whether to go further. If the numbers do not support a build, we say so.

Why the baseline is the whole point

The evidence on this is unusually blunt. MIT's Project NANDA study found 95% of enterprise GenAI pilots produced no measurable P&L return. S&P Global found the share of companies abandoning most of their AI initiatives rose to 42%, up from 17% a year earlier. Goldman Sachs found 76% of small businesses use AI, while only 14% have it fully embedded.

Read those together and the pattern is not that the technology fails. It is that almost nobody can say what changed, because almost nobody wrote down what things were like beforehand. An assessment that produces enthusiasm and no baseline is how you end up in the 95%.

What it costs

From $2,500, one to two weeks. The exact figure depends on how many workflows are in scope, and it is fixed in writing before the work starts — no hourly billing, no scope creep, no invoice you did not expect. If you go ahead with the 90-Day Implementation Sprint, the diagnostic fee is credited against it in full — so the diagnostic only ever costs money when it tells you not to build.

If it leads to a 90-Day Implementation Sprint, that starts from $15,000 and is quoted the same way: fixed, and agreed before anything begins.

When this is not worth buying

We would rather say this now than after the invoice:

• Under about ten people. There usually is not a workflow expensive enough for the maths to work.
• You want a proof of concept for a board. This produces a measured baseline, which is a duller and more useful object.
• Nobody senior owns the change. A baseline nobody is accountable for is a document, not a project.
• You already know the number. If you can already tell us what the workflow costs and it is signed off internally, skip this and go straight to a Sprint.

Who does the work

Senior practitioners. The founder leads every engagement, with senior developers and architects drawn from a pool of more than ten as the scope needs — not a team you were never introduced to on the call. Everything is built in your own accounts, so there is nothing to unpick if you stop working with us.

Questions people ask before booking one

Is this the same as an AI audit?

Same category, different ending. Most of what an AI audit for business produces is a report on what you could theoretically do — a list of opportunities, ranked. This produces one workflow, one metric, and that metric's measured value today, signed off by you. It is narrower on purpose, because a ranked list of twenty possibilities is not something you can act on or check later.

Why only one metric?

Because a project that is accountable for twelve numbers is accountable for none. One number, agreed up front, is what makes the result arguable in your favour later — or honestly against it.

What if you find nothing worth building?

Then we tell you, and that is the deliverable. It is a cheaper answer than discovering it after a $15,000 build, and it is the reason this is sold separately rather than bundled into a bigger engagement.

Can we skip it and go straight to a build?

Yes — if you already know which workflow, what it costs today, and that number is signed off internally. If any of those three is missing, the build has nothing to prove itself against.

How long does it take?

One to two weeks, depending on how many workflows are in scope. The price is fixed in writing before it starts.