Reinventors. How to make new money from AI, not just spend less.
Efficiency is a decision a CFO can model in an afternoon. New revenue from AI is a portfolio of bets, and it dies in the business case when it is valued the same way. This tool helps you find the play, test the riskiest assumption first, and fund it in stages a CFO can actually approve. It is how a company moves out of the Efficiency Trap and becomes a Reinventor.
Four revenue patterns · a six-axis lens · a four-stage funding ladder · no email required · download the one-pager
Four patterns, not a hundred.
AI makes new money in four patterns. Every candidate should map to one. If it does not, it is usually efficiency wearing a revenue costume.
Productised expertise
A service that was uneconomic to deliver one customer at a time, now delivered at scale. The senior human in the loop becomes the exception, not the unit cost.
New product surface
A capability that could not exist before the model, embedded as a feature or product a customer will pay for on its own.
New pricing model
Usage or outcome-based pricing that AI makes measurable and defensible, capturing value you were leaving on the table under a seat licence.
New reachable segment
A market that was too small or too costly to serve until the cost of serving it collapsed. The long tail, smaller accounts, a new geography.
Find the assumption most likely to kill it.
Take one candidate. Read it on six axes, low, medium or high. You are not ranking it. You are finding the one assumption most likely to kill it, because that is what the first stage of funding is built to test.
How the read is scored: Low counts as 1, Medium as 3 and High as 5, so six axes give a total from 6 to 30. A total of 6 to 12 starts at Frame, 13 to 18 at Probe, 19 to 24 at Pilot and 25 to 30 at Scale. These are the same bands the Claude prompt and the whitepaper use. Weak demand, delivery or defensibility can hold a play below its band, because those are the assumptions a stage is funded to test.
Score all six axes to get the riskiest assumption and the funding stage to start at.
Buy the bet in stages. Fund learning, not faith.
You do not ask a CFO to approve the whole thing on faith. You buy it in stages, each priced as an option on the next. Spend rises only as uncertainty falls, and you can kill cheaply at any gate. Killing is the system working, not failing. Score your play above and the stage you should start at lights up.
Frame
Write the bet in one sentence. Name the riskiest assumption. Set the kill criterion now, while you are still honest.
- Gate
- The bet, the assumption and the kill line are written down.
- Size
- About three weeks, two people.
Probe
The cheapest test of the riskiest assumption, almost always demand. Time-boxed, micro-budget, no build beyond the minimum.
- Gate
- Evidence of genuine pull.
- Size
- About twelve weeks, three to five people.
- Kill
- No pull. Stop, cheaply.
Pilot
Real customers, real small revenue, real delivery. You are now testing whether it repeats and what the unit economics look like.
- Gate
- Repeatability and workable unit economics.
- Size
- About six months, seven to ten people.
- Kill
- It will not repeat, or the economics do not hold.
Scale
Fund growth only once the economics hold and the play is defensible against a competitor buying the same tools.
- Gate
- Durable margin plus a moat you can name.
- Size
- Funded to the full model, with a six-month evaluation at the front.
- Kill
- Margin erodes or the edge is not real.
Three governance choices make the mechanism work. Ring-fence a small reinvention budget that does not sit under the same ROI bar as efficiency. Run it as a portfolio and expect most Stage 1 probes to die, because a portfolio where nothing dies is not taking real bets. And give it one named owner with a quarterly review where advance-or-kill decisions are made at the gates.
Whitepaper and Claude prompt
Two artefacts to run this with your CFO or your board. The whitepaper explains the four patterns, the six axes, and the four-stage funding ladder in depth. The Claude prompt runs the diagnostic conversationally against your specific idea and drafts the Probe brief the CFO can approve.
Fund Learning, Not Faith
The framework in twenty pages. Four revenue patterns, the six-axis qualifying lens, the four-stage funding ladder with evidence gates and kill criteria, and a CFO memo template that fits on one page.
Download PDF→The Reinventors Diagnostic, run live
Paste the prompt into a Claude project. Classify your AI idea against the four patterns, score it one to five on six axes, get a funding-stage recommendation and a draft Probe brief for your CFO. Ten minutes to run.
Copy the prompt→The tool finds the play. Installing the discipline is the work.
Most companies stay in the Efficiency Trap because reinvention has no home in the operating model or the budget. If you would like Ortent to help you build the portfolio and the governance around it, most engagements start with a Sprint. Not sure where you sit? The AI Value Gap self-assessment places you on the efficiency-versus-revenue map.
This tool is part of the field guide AI and the Board: the questions a board should ask about AI, and the tools that help answer them.
Read the guideFAQs
What is the Reinventors Diagnostic?
A self-assessment for growth-stage leaders weighing an AI-enabled new-revenue idea. It classifies the idea against four revenue patterns, reads it on six axes, and places it on a four-stage funding ladder (Frame, Probe, Pilot, Scale) with evidence gates and kill criteria. On this page you read each axis low, medium or high and the stage to start at lights up. The Claude prompt version scores each axis one to five and drafts a Probe brief for the CFO. Free, and no email is captured.
Who is the Reinventors Diagnostic for?
CEOs, CFOs and PE operating partners deciding how much to fund an AI-enabled new-revenue play, and whether to fund it as an experiment or a growth bet. Also useful for boards pressure-testing an idea before it becomes a line item. Answer for one AI-enabled revenue idea at a time, not the aggregate roadmap.
How long does it take?
A few minutes on this page if the idea is scoped: read the six axes low, medium or high and the starting stage lights up. The Claude prompt version takes about ten minutes and scores each axis one to five. No email, no signup, and the on-page tool runs entirely in the browser.
Do I need to give an email address?
No. Nothing is captured. The diagnostic runs entirely in the browser. There is no signup, no lead form, and no gated result.
What are the four revenue patterns?
Productised expertise (a service that was uneconomic to deliver one customer at a time, now delivered at scale). New product surface (a capability that could not exist before the model, built in as a feature or product). New pricing model (usage or outcome-based pricing that AI makes measurable and defensible). New reachable segment (a market that was too small or too costly to serve until the cost of serving it fell). Each pattern has different unit economics, evidence gates and kill criteria.
What are the four funding stages?
Frame (define the hypothesis and the killable evidence), Probe (buy the smallest experiment that can disprove it), Pilot (real customers and real economics), Scale (fund to the operating model). Each stage carries its own gate: a Frame idea graduates to Probe only when the biggest risk is cleanly framed and the killable evidence is designed.
Is this an AI ideation tool?
No. It is a funding-discipline tool. It assumes the idea already exists and helps the board decide how much to commit and at which stage. If the goal is idea generation, this is the wrong tool. If the goal is stopping bad ideas earlier and funding good ones with evidence, it is the right one.
What do I do with the result?
Two things. Fund only the stage the diagnostic recommends, not the pattern's ambition. Then run the diagnostic with the CFO independently. The axes where their score is lower than yours are where the case has not landed with finance yet, and those are the gates to close before the next raise.
Clarity when it counts.
For a board, advisory or fractional engagement, or a working call on the next stage, send a note. Thirty minutes is usually enough to test fit.
Request a 30-minute intro call