One Time vs subscription pricing
Choose One-Time or Subscription With a One-Page Worksheet
Fill in delivery cost, ongoing value, support, and a real buyer's reaction before you pick one-time or subscription.
Published 2026-11-05.
Pick a price the way you would pick a feature: write down what it has to support, then show it to someone who has the problem. Use it to make the offer and its economics concrete enough to test.
Use one page and four answers. A blank box means you do not have a price yet. You have a private guess.
The worksheet
| Question | What to write | How it changes the choice |
|---|---|---|
| Delivery cost | Variable cost to deliver and support one purchase, or one active customer per billing period; keep initial build cost separate | Compare the price with the appropriate delivery or billing period; a subsidized trial does not establish sustainable economics |
| Ongoing value | What new value or service the buyer receives in month two, even when they keep earlier outputs | A subscription needs continuing value, such as fresh work, maintained access, or a service you keep providing |
| Support burden | What a reasonable customer will ask a human after paying, and how many of those asks you can take | Heavy hands-on help can belong in a higher one-time fee, or in a support offer you can actually staff |
| Willingness signal | What one real buyer pays now for the workaround, and the words they used when they saw your number | A number nobody in the job has reacted to is still a note to yourself |
Read the ongoing-value row first.
- You deliver a finished result and provide no ongoing service: a one-time price is a useful starting hypothesis.
- The customer receives continuing value and you keep providing a service: a subscription may fit. Earlier outputs can still remain theirs.
- A finished core sits next to ongoing work: consider a one-time core plus an optional recurring service. Explain the split beside the price.
Another company’s tiers are not evidence about yours. Write down what is delivered once, what continues each month, the cost of that work, and why someone would want it again. Test that specific offer with people who have the problem before choosing the billing model.
Test one clear offer with a small batch
- Fill the four rows in a document the buyer will not see. Ranges are fine. False precision is not. "About half a day of help per customer" is more honest than a made-up margin.
- Put one number and one offer where a buyer can read them. If they are meant to buy without a call, do not hide the number behind a contact form.
- Choose a small batch, such as five people who already have the problem. Ask what they pay now, in time or money, then show the specific offer. Record questions and objections. This is a planning batch, not a representative market sample.
- Ask for a concrete next step, such as a paid pilot with clear terms. Keep stated intent separate from an actual payment and continued use.
- Add a second tier only when those conversations reveal two different jobs. Three cards will not fix a muddy offer.
How to Add Stripe Payments to a Vibe-Coded App is the next practical step after the offer is clear. Checkout can charge once or on a schedule. The integration does not decide which one fits the worksheet, and installing Stripe is your change to approve.
If people cannot tell what they are buying, fix the offer before you debate the model. Landing page best practices is about what a visitor can see on the page. Pricing page examples show public patterns. A pattern does not justify invented savings, a "most popular" badge you do not mean, or customer counts you do not have.
Hypothetical example
**Hypothetical.** You sell a downloadable pack of data-cleaning templates. Delivery cost for another buyer is low once the files exist, and the buyer keeps them. A one-time offer may fit that finished result. If you later provide new templates every month or maintained import tooling, describe that continuing service separately and estimate its support cost. Test the complete offer with suitable buyers. A positive answer is stated intent; an actual paid pilot gives you a different kind of evidence.
Copyable agent prompt
Help me choose one-time, subscription, or a split, using only the worksheet I paste. Do not invent benchmark prices, conversion rates, or customer quotes. Do not give tax, legal, or accounting advice.
Restate each worksheet row in one line. If a row is blank, stop and ask for it.
Then propose:
- One model, tied to the ongoing-value row.
- One number I already supplied, or a blank I must fill. Do not create a price.
- The sentence beside the price, including what is not included.
- Three questions I will ask in a real conversation.
- A "do not publish yet" list of any claim in my draft that the worksheet does not support.
Keep the buyer's words separate from your suggestions.Finish rule
You are done when all four rows are filled, the page states one model and one number, and your notes include at least one person in the target job who saw that number and answered with a yes, a no, or a counteroffer. Changing the number later is normal. Publishing three tiers so you can avoid that conversation is not a finish.
A free scan can show whether the public page states the offer in a way a cold visitor can follow. It cannot choose the price.