Billing monthly does not make revenue recurring. If the problem is solved once, the customer leaves once it is solved, and the subscription you built becomes an acquisition treadmill with a nicer invoice. At $29 a month against $3,000 of target income, the difference between a problem that returns and one that does not is 17 times the sales work, forever. That is the number to settle first.
The variable is not the price
Choose a price and a target and the customer count falls out immediately — division, not strategy. Our free calculator does it, and so does the product's paid pricing analysis; they are separate functions held to the same arithmetic by a test that fails if they ever disagree. At $29 a month you need 104 paying customers to reach $3,000.
That count is the easy half and it is not what decides a subscription. What decides it is how many of those 104 are still there next month, because whoever is not there has to be replaced before you grow at all. Two businesses with the same price and the same target can differ by an order of magnitude in how much selling they require, and nothing on the pricing page shows it.
There is one place the arithmetic says so out loud. Ask the calculator how many customers a one-time price needs to reach a monthly target and it returns nothing — not a zero, not an estimate. The question is malformed, and returning a number would be the polite version of a lie.
What standing still costs, at each level of repeat
Hold the price and the target still. Sweep how much of the base leaves each month:
New customers required every month purely to hold $3,000, at $29 a month. Counts come from the same function behind the customers-needed calculator, so the chart cannot disagree with the tool. Bars are relative to the largest.
Read the two ends against each other. A product people stay with needs 6 new customers a month — 72 a year, and every one you find stays found. A product that is finished when the job is done needs 104 a month, 1248 a year, and none of them count twice. Same price, same income, 17 times the work.
That is why "we will just charge monthly" is not a pricing decision. It is a decision to run the harder business without telling yourself you have chosen it.
The three shapes, and what distinguishes them
Most ideas are not at either end. They are somewhere in the middle, and the middle has its own correct answer:
| Shape | After the job is done | What the complaints look like | What to charge |
|---|---|---|---|
| Solved once | The buyer is done and has no reason to open it again | Complaints describe a project with an end date | One-time price, or a service |
| Comes back rarely | Needed again in months or years, unpredictably | The same people complain again, far apart | Pay per use, or a low keep-the-lights-on tier |
| Never finishes | The work continues and the product keeps doing it | Complaints describe a routine, not an event | Monthly, and the number above holds |
A classification, published so it can be argued with — the boundaries are ours, not a measurement. The column that does the work is the third: it is the only one you can check without customers.
The middle row is where the money usually is and where founders least like to sit, because pay-per- use looks smaller on a spreadsheet than a subscription does. It is also the only one of the three that survives contact with a buyer who genuinely needs the thing twice a year.
How to tell which one you have, before you build the billing
None of this requires customers, and none of it requires a call.
- Read the complaints for an end date. Go back to the posts and threads that made you believe in the problem. Does the person describe a project — a migration, a launch, a filing, a cleanup — or a routine they will be doing again next Tuesday? A project has a last day. Write down which one you are actually reading, not which one you would prefer.
- Look at what they use now. If the current alternative is a spreadsheet they rebuild every month, the problem recurs. If it is a spreadsheet they built once and still have, it does not.
- Sell the one-off version first. Charge for the single job at a real price, async, with no pilot and no call. If enough of the same buyers come back inside a quarter, you have found the recurrence, and you now have the evidence to price a subscription. If they do not come back, you have learned it for the cost of a landing page instead of the cost of a billing system.
- Then ask the two published questions. Our single-idea check runs every idea past the same fixed list, and two of those filters are exactly this: Can it charge on repeat? and Pays for itself within a year?. An idea that cannot answer the first is not a subscription, whatever the pricing page says.
What the arithmetic cannot settle
It cannot tell you your churn before you have customers. Every number in the table above is a probe, not a forecast, and the useful reading is the ratio between the rows rather than any single row.
It also cannot rescue a genuinely one-time product, and it is worth being clear that plenty of one-time products are good businesses. The failure mode is not selling something people buy once — it is building the cost structure of a subscription business on top of it and then being surprised by the sales load. Pick the shape deliberately and the arithmetic is on your side either way.
The short version
- The customer count comes from price and target. The amount of selling comes from how many of them stay, and nothing on your pricing page shows that.
- At $29 and $3,000, a product people stay with needs 72 new customers a year. Solved-once needs 1248.
- A one-time price has no monthly-income answer at all — our calculator refuses the question rather than returning a zero.
- The check is in the complaints you already collected: a project has an end date, a routine does not. Sell the one-off first, then price the repeat.