A listing tells you the revenue and the price. Divide one by the other and you have the number the advert left out — how many customers that revenue implies — and from there, how many hours a week they would cost. At $2,400 a month, two listings can differ by 23 times in implied customer count and by a whole support band. Same revenue, different job.
The arithmetic the advert leaves to you
Here is one monthly revenue figure reached three ways. The only thing that changes is the price:
| Price | Customers behind the revenue | Support a week | Band |
|---|---|---|---|
| $8 | 300 | 17.3 h | heavy |
| $29 | 83 | 4.8 h | moderate |
| $199 | 13 | 0.7 h | light |
All three rows are $2,400 a month at 15 support-minutes per customer. Counts and hours come from the same two functions behind customers-needed and support hours, so the table cannot disagree with the tools. Read the last column downwards: it is the job you are buying.
The top row is a business with 300 customers and 17.3 hours of support a week before you have improved anything. The bottom row is 13 customers and under an hour. Identical revenue, and one of them is somebody's evening job.
That is the first question to ask about any listing, and you can put a number on it from two figures already in the advert before the seller says anything. Then ask which direction it moves: a cheap, high-count business gets harder as it grows, and the growth is what you are paying a multiple for.
One thing that division assumes, and it is worth saying out loud. Revenue over list price is the real customer count only if every customer pays that one price — no legacy plans, no annual discount, no coupons, no second tier. Few running businesses clear that bar, so treat the figure as what the advert implies rather than what the business is, and ask the seller for the paying-account count early. If the two numbers disagree, you have learned something better than the count: there is a plan mix, and a plan mix moves the support shape as much as the headcount does.
Fewer of the idea questions are settled than you would think
If you have read anything on this site about checking an idea, the instinct is to run the same list and expect the trading history to have answered it. It answers less than it looks like, because a running business is evidence about the past and half of the list is a question about the future:
| The check | On a running business | What to ask for |
|---|---|---|
| Can it make money? | The numbers answer it | Revenue exists. Ask for the processor export, not a screenshot |
| Can it charge on repeat? | The numbers answer it | Recurring or not is visible in the same export |
| Can it charge enough? | The numbers answer it | Revenue divided by paying accounts is the real average price |
| Can you reach buyers? | Still open after the sale | The seller's channel may be a person, a network or an algorithm you do not inherit |
| Low hands-on support? | Still open after the sale | Ask for a month of the support inbox, counted |
| Safe from one platform's control? | Still open after the sale | Which platform sends the traffic, and which one the product plugs into |
| A real product, not just an AI wrapper? | Still open after the sale | Revenue proves people buy it, not that it outlives the thing it wraps |
| Can you sell and run it? | It is about you | Nothing in the listing can answer this, and it decides the outcome |
| No heavy compliance blocker? | Still open after the sale | Unenforced is not compliant. Ask what regime it trades under and who decided that |
| Sells without 'book a call'? | The numbers answer it | Look at how the last ten customers actually signed up |
| Pays for itself within a year? | The numbers answer it | Asking price against profit, not against revenue |
| Can it be sold or handed off? | Still open after the sale | Changing hands is not the same as running without the previous pair |
The twelve checks a single-idea report runs, in its own words, with our judgement of where each one stands once money is already changing hands. 5 are settled by the numbers, 6 survive the sale, and 1 was never about the business at all.
The 6 that survive are the ones worth your diligence time, and they share a shape: each is a question about whether the thing that works today keeps working when the owner changes or the ground under it does. Three of them look settled and are not — a business that has traded unchallenged has not been ruled compliant, revenue does not prove a product outlives what it wraps, and an asset changing hands is not the same as an asset that runs without the previous pair of hands.
The three questions a listing never asks itself
None of these appear on an idea checklist, because an idea has no customers to concentrate and no owner to depend on.
- Revenue concentration. What share of the money comes from the largest account, and from the largest five. A business where one customer is a fifth of revenue is a business with a single point of failure you are buying at a multiple.
- Owner dependence. Which parts of the operation are a person. The support inbox answered in a particular voice, the one integration that breaks monthly, the affiliate who sends half the traffic and is friends with the seller. Ask what happens to each on the day the seller stops caring, because that day is the day after closing.
- Where the customers came from, and whether that door is still open. A product built on a channel that no longer works is a product with a slowly shrinking customer base and a very good trailing twelve months. This is the check that most often separates a fair price from a bad one.
Two documents settle more of this than any conversation: the payment processor's own export, and a month of the support inbox. Both are refusable, and a refusal is itself an answer.
What this does not settle
None of it tells you the business will keep growing. The arithmetic above bounds the work and the shape; it says nothing about whether demand is expanding or the market is being eaten by something adjacent.
It also cannot price the thing. Multiples are a market question, and this page has no opinion on them — its argument is only that two listings quoting the same revenue can imply different amounts of work, and that the difference shows up before you ask the seller anything.
And the check that matters most is the one no listing can help with: whether you, specifically, will sell and operate this after the excitement wears off. That question has the same honest answer it has for a new idea.
The short version
- Revenue divided by price is the customer count the advert implies, and that count times your support estimate is the job. Both come out of the advert — and both are exact only if everyone pays one price, so ask for the paying-account count and treat a gap as a finding.
- At one revenue figure, the cheap and dear versions differ by 23 times in implied customers and by two whole support bands.
- Of the 12 questions a single-idea check asks, only 5 are settled by a business that is running. Spend the diligence on the 6 that are not.
- Concentration, owner dependence and the channel's health are the acquisition-only questions. Ask for the processor export and a month of the support inbox.