Free guide · 47 pages
Could your expertise become an AI product?
Some of the strongest AI product opportunities don't start with a new idea. They start with something a business already has: specialist expertise, a proven process, recurring demand it can't profitably serve, or data nobody else could recreate.
This guide gives you the fifteen-box checklist for spotting that, then the workbook for deciding whether it is worth building — scoring the opportunity, auditing your foundations, a 30-day validation plan, what a build really costs to run and price, and the conditions under which you should stop.
PDF, 1.2 MB. No email required — you already gave us those details.
Five signs, fifteen boxes
Tick what is true of your business. The guide tells you what each score means, and which sign to act on first.
Your expertise is valuable, but difficult to scale
You turn away work that's too small to service profitably
Clients ask for your framework, not just the outcome
Your market is beginning to productise around you
You have the data nobody else can create
Fourteen parts, written to be worked through
Each part ends with something to fill in, decide or check inside your own business. None of it requires a technical brief.
- Score yourselfCount your ticks, and find the sign worth acting on.
- Why "could we?" is the wrong first questionWhat the UK adoption data actually says about why AI projects fail.
- The three shapes a product can takeInternal tool, client add-on, standalone product — and the rule for choosing.
- Pick one: the opportunity scorecardEight criteria, scored 0–2, including the 10x model test.
- The foundations auditProcess, data, contracts, personal data and ownership — each has killed products after the money was spent.
- Validate before you build: a 30-day planTwenty conversations, a commitment ladder and pricing without asking "what would you pay".
- The moneyWhat it costs to find out, to build, to run, and how to price it.
- How a build actually runsFour stages, and the questions to ask anyone who wants to build it for you.
- What AI can and can't do reliablyHallucination rates, human baselines, and the evaluations to insist on.
- Regulation as a product signalThe current UK list, and how a new compliance regime becomes a market.
- What changes inside the businessOwnership, professional-body rules, insurance, and the numbers to watch.
- Kill criteriaHow to know when to stop — decided before you start.
- A worked example, start to finish
- Where to go next
The next step isn't to build it
It's to find out whether it's worth building — and that's a four-week piece of work, not a four-month one. The AI Opportunity Finder takes about ten minutes and turns your answers into a first view of where the opportunity sits.