Most AI spending is a guess. Yours does not have to be.
Answer twelve questions about how your business actually runs. Get back a scored, priced audit of where AI is worth it, what each change gives you back in hours and money, and what to leave alone.
Same day, no call required. Every number comes from a figure you gave us, and the sums are printed next to the answers so you or your bookkeeper can check them.
Someone signs up for something promising, three people try it, and it quietly lapses. Nobody agreed what it was meant to improve, so nobody can say whether it did.
Weeks later you have a deck about industry trends. It is not wrong. It also does not tell you which job to change on Monday, or what changing it is worth to you.
The most common outcome by far. Everyone agrees AI matters, nobody can point to a first step they would defend out loud, and the year passes with the same people doing the same work by hand.
All three go wrong the same way. Nobody wrote down what the work costs today, so nobody can ever prove whether it got better.
A report you can argue with, line by line.
It opens with one recommendation, the hours it gives back, what that is worth each month, and what a year of waiting costs you. Then it shows the sums.
Six to eight opportunities, each scored, each with a setup cost and how long it takes to pay for itself. Nothing is called transformative. Things are called worth doing, worth trying small, or not worth it yet.
And it names what you should skip, with the reason for each. That is usually the part people bring up first.
Not a quiz, and it is the hour you are paying for. You list the jobs that eat the most time, who does them, and what an hour of that person’s time costs you. The money side is built from your figures, not from industry averages.
Your answers are scored against the eight things that decide whether an AI project survives contact with a real business, then priced against what the tools cost today. The report is yours straight away, ready to forward or save as a PDF.
It ends with two or three things to measure before you build anything, and a space to write today’s number. That is what lets you prove a year from now whether it worked.
You give one figure: what an hour of your team’s time really costs, salary plus overhead. Every dollar in the report is that figure times the hours saved, minus what the tools cost. The sum sits next to the answer, so you can check it or tell me the figure is wrong.
Each idea is scored on how much it would help, how doable it is, how complicated and expensive, how ready you actually are, and what could go wrong. The low scores are the point. If everything wins, the ranking tells you nothing about what to do first.
The report names what to measure before you build, how to capture each one in under an hour, and what good looks like afterwards. Skip it and you will never be able to show the change paid for itself, which is how most AI spending ends up impossible to judge.
What you already pay for, how organised your files are, what you can spend a month, and how your team handled the last new tool all limit what gets recommended. The plan is capped by what you will actually use, not by what is possible in theory.
It will tell you what to skip.
Every audit names the ideas that came up and did not make the cut, and gives the specific reason each one was ruled out: what you can spend, the state of your files, or how your team has handled new tools before.
A report that recommends everything is a sales pitch. This one is built to be argued with, which is the only reason to trust the parts it does recommend.
Ange Ingabire
I run the AI Productization Lab, where I build and ship AI tools as working products rather than pilots. That means I spend my time on the unglamorous half of this work: what integrates with what a business already runs, what a team will actually adopt, and what quietly gets abandoned three weeks after launch.
I built this audit because I kept watching capable businesses spend real money on AI without anyone establishing what the work cost them beforehand. Every engagement I have taken on since has confirmed the same thing: the constraint is almost never the technology, and the opportunities worth doing are usually smaller and duller than the ones people arrive excited about.
The scoring. Every idea is judged on the same eight things, weighted the same way, with the same cut-off deciding what gets recommended and what gets ruled out. That sameness is the point: it means you are comparing one opportunity against another, not against how enthusiastic someone sounded, and it is why the bad scores get printed next to the good ones.
The writing and the research are model-assisted, and the tool prices are checked at the time of writing rather than remembered. The money is not: it is worked out from the hours and the hourly cost you give, which is why the sums are printed instead of just the answers. The scoring, the eight things each idea is judged on, and the rules that decide what gets ruled out are mine, applied the same way every time. That last part is what makes two audits comparable.
No. The fee is non-refundable, because the work is delivered in full the moment you receive it and there is nothing to take back. That is also why the entire sample report is published on this page rather than teased: you can read a complete audit end to end, judge the depth and the scoring, and decide before you pay rather than after.
Then you have paid a small amount to avoid spending a large amount, and you have the reasoning in writing for the next time someone pitches you. It happens. If your total time load comes back too small to justify any change, the report says so in the first paragraph rather than burying it.
Sometimes the answer is obvious and nobody has ever priced it. Knowing that writing proposals costs you a specific amount every month, and that fixing it pays for itself in a specific number of months, is different from having a hunch about it. An obvious answer with a number attached is what gets a decision made.
Small is where this works best, because one person’s time is a big share of what you can deliver and the jobs are easy to see. Large companies have the opposite problem. If you are two people, the answer may well be that your time is better spent selling than automating, and the report will tell you that.
No, and it is better if you do not arrive with a shortlist. The questions are about how your business runs: who does what, how often, where things get stuck, what an hour costs you. Nothing asks you to have an opinion about tools.
The full report, same day, as a document you can forward, print or save as a PDF. It is written to be read by someone who was not in the room, because that is usually the person who controls the money.
Up to you. The audit is the whole product. The ninety day plan is written so your own team or the contractors you already use can run it without me, and there is no implementation package waiting at the end of it.
One hour of your time, and you will know where to start.
Billed at $600 an hour, and the audit takes one hour, so $600 in total. There is no call to book and nothing to sit through: you spend the hour answering twelve questions when it suits you, and the report is yours the same day. The full sample report is published in advance, so you can judge the format, the depth and the scoring before you pay anything.
The audit is an advisory opinion, not a guarantee of results. All hours and costs in it are estimates supplied by you and are not independently verified, so the figures it produces are projections rather than promises. Nothing in the report constitutes financial, legal, tax or employment advice, and you should take professional advice before acting on anything with contractual, regulatory or staffing consequences. Tool prices and capabilities are checked at the time of writing and change frequently. Analysis is produced with the assistance of AI systems under human review, and your answers are processed accordingly, so do not submit personal data, client confidential material or anything covered by an NDA. Any outcome depends on how the recommendations are implemented and adopted, which is outside my control. The fee is non-refundable and buys the audit itself, not a result.