Capability template for leadership
Business model assumptions review
Before you commit money to a new or revised business model, see it block by block with every claim marked evidenced or assumed, the riskiest assumptions ranked, and the cheapest test for each. You choose what to test; no projection or valuation is made.
Listed under
Add it to a workflow in the Fully Powered Planner, beside the teams and roles that will run it. Book a fit call
The outcome
The sponsor sees the whole model on one page and can tell, claim by claim, what is known and what is hoped. The page lays out six blocks (customers, offer, channels, revenue, costs and partners), counts how much of each is backed by something that happened, and ranks the assumptions by how badly the model suffers if they are wrong and how unsure the sponsor is. For the riskiest few it names the cheapest way to find out.
A claim counts as evidenced only when money changed hands, people used the thing, or someone measured it, and the evidence has a date. A customer saying they like the idea, or an estimate, keeps the claim in the assumed column. That keeps a warm conversation from passing for proof.
Nothing is decided for you. The review ends with a choice line for each of the riskiest assumptions: test now, later, or accept as it is. It makes no sales forecast, margin figure or valuation, and it sets no pass or fail line for you.
Who starts the work, and when
The founder, or the executive sponsor of a new line of business, starts it before spending serious money or time on a model: a new offer to the market, a change of price structure, a move from one-off sales to a subscription, a new customer group. It is worth running again whenever the model changes or a test comes back.
The sponsor writes a short brief and fills in a list of claims, one row each, with the evidence behind it and two ratings. The AI assistant lays out and ranks; the sponsor rates, chooses which assumptions to invest in testing and sets what result would count as a pass.
What they supply
- A short brief. The model and what is changing, who sponsors it, the decision the review serves and when, the review date, and any context worth knowing.
- The claims list. One row per claim, filed under customers, offer, channels, revenue, costs or partners. Each row gives the kind of evidence (paid, used, measured, said, estimate or none), the evidence in a line, its date, and, for what is assumed, two ratings: how much the model suffers if the claim is wrong, and how unsure you are.
- Test ideas, if you have them. A test you already have in mind for a claim, with its cost and the days it takes.
What comes out
One page in a fixed order. The model block by block, each claim marked evidenced or assumed; how many claims in each block are evidenced; the assumed claims in rank order; for the riskiest, the cheapest test, either yours or a proposal from a fixed ladder, with the pass and fail lines left empty for you; the evidenced claims whose proof has grown old; the questions only the sponsor can answer; and a choice line for each risky assumption.
Where something is missing, such as a block with no claims or an assumption nobody rated, the page says so and asks. It never fills the gap itself.
An invented example
Illustrative example with invented data. The company, its customers, its people and every date and figure below are invented. "Company A" is an invented regional business that sold office plants one order at a time and is deciding whether to move to a monthly plant-care subscription. People appear by role only. First come the inputs the sponsor supplies (the brief and the claims list), then the review.
The inputs
1. The brief (written by the sponsor)
Model: Monthly plant-care subscription for small offices, revised from one-off plant sales
What is changing: from selling plants once to a monthly price that includes a visit to keep them healthy
Sponsor: Founder
Decision this serves: Whether to move the whole sales effort to the subscription from January 2027
Decision date: Friday 11 December 2026
Review date: Tuesday 10 November 2026
Context: Company A had 60 one-off plant orders from 2025-11 to 2026-10, and 12 offices now pay for occasional care visits. 9 office managers were asked about the subscription in October 2026.
Settings: defaults
2. The claims list (typed by the sponsor, one claim per row)
block,claim,evidence_kind,evidence,evidence_date,if_wrong,how_sure,test_idea,test_cost,test_days
customers,"Offices of 10 to 50 staff with someone who orders for the office are the buyers.",used,"41 of the 60 one-off orders last year came from offices of that size.",2026-09,,,,,
customers,"Office managers will pay monthly rather than per order.",said,"6 of the 9 office managers asked said monthly billing suits them.",2026-10,3,2,,,
offer,"Offices want the plants kept alive, not only supplied.",said,"4 of the 9 office managers asked said dead plants were why they stopped buying.",2026-10,3,1,,,
offer,"One visit of 45 minutes a month keeps a 12-plant office healthy.",,,,2,3,"Keep 3 of the 12 current care offices on a monthly 45-minute visit for 3 months and photograph the plants each month",180,90
channels,"Building managers keep referring tenants.",used,"Two building managers sent 7 orders last year.",2026-06,,,,,
channels,"Referrals and the web form will bring 5 new subscribing offices a month.",estimate,"Founder's estimate: 60 orders last year is 5 a month.",2026-10,3,3,,,
revenue,"Offices will pay $85 a month for the 12-plant service.",,,,3,3,"Offer 3 current customers a written 3-month order at $85 a month",120,21
revenue,"Subscribers stay at least 12 months.",said,"2 of the 9 office managers asked said they would stay a year.",2026-10,3,3,"Look up how long the 12 current care offices have stayed",0,2
costs,"A visit takes no more than 45 minutes including travel.",measured,"Timed 6 visits in September; the longest was 45 minutes.",2026-09,,,,,
costs,"Replacing dead plants costs under 4% of subscription revenue.",,,,2,3,,,
costs,"Van cost is $9 a visit.",measured,"August 2025 fuel receipts divided by 31 visits.",2025-08,,,,,
partners,"The wholesale grower will hold its prices for a year.",said,"The grower's sales rep said so on a call.",2026-10,2,2,,,
partners,"A second grower could supply us if the first fails.",,,,2,,,,
The review
BUSINESS MODEL ASSUMPTIONS REVIEW: Monthly plant-care subscription for small offices
Model: Monthly plant-care subscription for small offices, revised from one-off plant sales; from selling plants once to a monthly price that includes a visit to keep them healthy.
Sponsor: Founder. Decision this serves: Whether to move the whole sales effort to the subscription from January 2027. Decision date: Friday 11 December 2026.
Context: Company A had 60 one-off plant orders from 2025-11 to 2026-10, and 12 offices now pay for occasional care visits. 9 office managers were asked about the subscription in October 2026.
Review date: Tuesday 10 November 2026. Prepared from the brief and the claims list; default settings. The review makes no projection or valuation.
THE MODEL, BLOCK BY BLOCK
CUSTOMERS
C01 EVIDENCED (used, 2026-09). Offices of 10 to 50 staff with someone who orders for the office are the buyers. Evidence: 41 of the 60 one-off orders last year came from offices of that size.
C02 ASSUMED (said). Office managers will pay monthly rather than per order. Evidence: 6 of the 9 office managers asked said monthly billing suits them.
OFFER
C03 ASSUMED (said). Offices want the plants kept alive, not only supplied. Evidence: 4 of the 9 office managers asked said dead plants were why they stopped buying.
C04 ASSUMED (no evidence). One visit of 45 minutes a month keeps a 12-plant office healthy. Evidence: none.
CHANNELS
C05 EVIDENCED (used, 2026-06). Building managers keep referring tenants. Evidence: Two building managers sent 7 orders last year.
C06 ASSUMED (estimate). Referrals and the web form will bring 5 new subscribing offices a month. Evidence: Founder's estimate: 60 orders last year is 5 a month.
REVENUE
C07 ASSUMED (no evidence). Offices will pay $85 a month for the 12-plant service. Evidence: none.
C08 ASSUMED (said). Subscribers stay at least 12 months. Evidence: 2 of the 9 office managers asked said they would stay a year.
COSTS
C09 EVIDENCED (measured, 2026-09). A visit takes no more than 45 minutes including travel. Evidence: Timed 6 visits in September; the longest was 45 minutes.
C10 ASSUMED (no evidence). Replacing dead plants costs under 4% of subscription revenue. Evidence: none.
C11 EVIDENCED (measured, 2025-08). Van cost is $9 a visit. Evidence: August 2025 fuel receipts divided by 31 visits.
PARTNERS
C12 ASSUMED (said). The wholesale grower will hold its prices for a year. Evidence: The grower's sales rep said so on a call.
C13 ASSUMED (no evidence). A second grower could supply us if the first fails. Evidence: none.
EVIDENCE BY BLOCK
Customers: 1 of 2 evidenced.
Offer: 0 of 2 evidenced. All assumed.
Channels: 1 of 2 evidenced.
Revenue: 0 of 2 evidenced. All assumed.
Costs: 2 of 3 evidenced.
Partners: 0 of 2 evidenced. All assumed.
Total: 4 of 13 evidenced, 9 assumed.
RANKED ASSUMPTIONS
1. C08 Revenue. If wrong 3, how sure 3, score 9. RISKIEST. Subscribers stay at least 12 months.
2. C07 Revenue. If wrong 3, how sure 3, score 9. RISKIEST. Offices will pay $85 a month for the 12-plant service.
3. C06 Channels. If wrong 3, how sure 3, score 9. RISKIEST. Referrals and the web form will bring 5 new subscribing offices a month.
4. C02 Customers. If wrong 3, how sure 2, score 6. RISKIEST. Office managers will pay monthly rather than per order.
5. C04 Offer. If wrong 2, how sure 3, score 6. RISKIEST. One visit of 45 minutes a month keeps a 12-plant office healthy.
6. C10 Costs. If wrong 2, how sure 3, score 6. Replacing dead plants costs under 4% of subscription revenue.
7. C12 Partners. If wrong 2, how sure 2, score 4. The wholesale grower will hold its prices for a year.
8. C03 Offer. If wrong 3, how sure 1, score 3. Offices want the plants kept alive, not only supplied.
Not yet rated: C13 (how sure is missing).
Also at or above the risk score, beyond the first 5: C10 (score 6).
CHEAPEST EVIDENCE FOR THE RISKIEST
1. C08 Subscribers stay at least 12 months.
Test: Look up how long the 12 current care offices have stayed. Cost: 0. Days: 2.
Pass line: sponsor to set. Fail line: sponsor to set.
2. C07 Offices will pay $85 a month for the 12-plant service.
Test: Offer 3 current customers a written 3-month order at $85 a month. Cost: 120. Days: 21.
Pass line: sponsor to set. Fail line: sponsor to set.
3. C06 Referrals and the web form will bring 5 new subscribing offices a month.
Test: Proposed, rung 4: run one small, capped trial with a real customer or channel. Cost: not stated. Days: not stated.
Result would be: used or paid
Pass line: sponsor to set. Fail line: sponsor to set.
4. C02 Office managers will pay monthly rather than per order.
Test: Proposed, rung 3: ask for a small written commitment, such as a deposit or an order. Cost: not stated. Days: not stated.
Result would be: paid
Pass line: sponsor to set. Fail line: sponsor to set.
5. C04 One visit of 45 minutes a month keeps a 12-plant office healthy.
Test: Keep 3 of the 12 current care offices on a monthly 45-minute visit for 3 months and photograph the plants each month. Cost: 180. Days: 90.
Pass line: sponsor to set. Fail line: sponsor to set.
A paid, used or measured result moves a claim to evidenced.
A result that someone said does not.
EVIDENCED CLAIMS TO RECHECK
C11 Van cost is $9 a visit: evidence dated 2025-08, 15 months before the review date; the limit is 12 months.
OPEN QUESTIONS FOR THE FOUNDER
1. C11 Costs: the evidence is dated 2025-08, 15 months before the review date, beyond the 12-month limit. Is it still true, or does it need a fresh measurement?
2. C13 Partners: you rated the harm if wrong as 2 but gave no rating for how sure you are, so it is not ranked. How sure are you: 1 fairly sure, 2 leaning, 3 a guess?
3. Pass and fail lines: none are set for C08, C07, C06, C02 and C04. For each, what result would count as a pass and what as a fail, before any test starts?
SPONSOR'S CHOICE
C08 [ ] Test now [ ] Later [ ] Accept as is
C07 [ ] Test now [ ] Later [ ] Accept as is
C06 [ ] Test now [ ] Later [ ] Accept as is
C02 [ ] Test now [ ] Later [ ] Accept as is
C04 [ ] Test now [ ] Later [ ] Accept as is
Prepared for the Founder's decision. Nothing has been decided, sent, ordered, booked or changed. No projection or valuation has been made.
The founder reads the review, answers the three questions and ticks one box for each risky assumption. The review only prepared it.
Where it fits
- Leadership, for owners and sponsors deciding where to put the next dollar.
- Product management, when a product or price structure is changing and the team needs to know which parts are still guesses.
- Professional services, when a firm is turning custom work into a packaged or recurring service.
- Software and technology, when a new product or pricing model is about to be built on assumptions about buyers.
- Ecommerce and retail, when a store adds a subscription, a new channel or a new customer group.
- Consumer brands, when a brand tests a new product line or a new way to sell.
- Startups and new ventures, where almost everything in the model is still an assumption.
- New business lines, before a team is moved onto the new line.
- Growth & scaling, when a model that worked at one size is stretched to another.
What has to change for your company
Decide who sponsors: the person who decides where the money goes. Keep or change the two rating scales and the risk score at which an assumption counts as risky (6 by default), how many are shown as the riskiest (five), and how old evidence may be before it is rechecked (twelve months). Decide which kinds of evidence count; the default is paid, used or measured. Add your own default test for each block if the ladder does not suit how you sell. It runs in ChatGPT, Claude, Copilot or Gemini from the brief and the list you type or paste; it needs no connection to any system.
What has been checked
Fully Powered AI wrote this capability template, drawing on a published open-source package credited in its sources. It was editorially reviewed on 10 November 2026. It was tried twice on its invented example in Claude on 10 November 2026, the second time after the instructions were revised, and the record of the second run is published with the template. It has not been tested in a client environment or as a packaged version in ChatGPT, Claude, Copilot or Gemini.
Testing in your environment means running it on a model you already know the outcome of, with the claims written as they stood at the time, plus hard cases written in advance: a block with no claims, a rating outside the scale, evidence with no date, a claim that says "customers love it" with no kind. The sponsor checks the ranking against the ratings before relying on it. Offer review takes the offer block further, and Premortem before a decision is the next step once the sponsor has chosen what to commit to.
The full capability template
Everything above is the outline. The complete capability template is the working version: the roles, the inputs, the procedure, the exceptions and stop conditions, the output format, the reviewer checklist, the acceptance tests, the instructions for an AI assistant and what to customize for your company. It comes with a download and the link to its folder in Fully Powered’s public repository.
Free with an account. Sign in by email: no password and no card.
Next step
Put this capability to work for your team: on your plan, or built with you in a sprint.
Build this in a sprint
In a sprint, Fully Powered adapts this capability template to your rules, proves it on your own work and puts it in the hands of the roles that own it.
Two ways to begin: