On this page
Cost-plus does not work here. The second copy of a file costs nothing to make, so "cost plus a margin" gives you a number with no information in it. What is left are three anchors that do carry information, none of which require a single sale to use.
And one piece of mechanics worth knowing before you choose: the fixed fee on every card payment means very cheap products lose a surprising share of their price. What taking your first payment online actually costs has the arithmetic.
What are the three anchors?
What would they otherwise do? Not "what is the competing product" — what is the actual alternative, which is usually doing nothing, or spending a weekend working it out from free material. If your thing saves a competent person six hours, that is the comparison, and six hours is worth something you can name.
What does the problem cost them? A guide that stops someone buying three overlapping subscriptions saves them a few hundred euros a year. That does not mean you can charge a few hundred euros, but it sets the ceiling the price has to sit comfortably below.
What do comparable things sell for? Not to copy, but to know where the expected range is. A price outside the range needs a reason the buyer can see in the first ten seconds.
| Anchor | What it gives you |
|---|---|
| The alternative | a floor — below this, why would they bother |
| The cost of the problem | a ceiling — above this, the maths stops working |
| The category | the range buyers expect without explanation |
Three numbers, usually in that order of reliability. If they disagree wildly, the thing you are selling is probably two products.
Which direction is easier to move?
Down.
Raising a price is unpleasant: the people who bought at the old one feel it, and you will be tempted to grandfather them, which complicates everything. Lowering is easy, and a discount is also a reason to send an email, which is a second thing you needed.
So if you are genuinely unsure, pick the higher of two candidate numbers. The cost of being 20% too expensive for three months is some lost sales. The cost of being 50% too cheap for three months is lost sales and a repricing conversation.
The Short Version
Every conclusion from all eight playbooks — 97 chapters, condensed. Free.
Is cheap ever the wrong answer?
Below about ten euros, usually.
Two reasons. The mechanical one: a fixed fee of €0.25 per transaction is 1.25% of a €20 sale and 8.3% of a €3 one, so the economics of very small prices are worse than they look.
The other is that price is a signal. A thing priced at €3 is telling the buyer something about how much work went into it, whether they will get support, and whether it will still exist next year. Sometimes that signal is accurate and fine. Often it is the reason a serious buyer scrolls past.
How do you find out whether the number is wrong?
Watch where people stop, not whether they buy.
| Where they stop | What it usually means |
|---|---|
| Never reach the product page | a message problem, not a price problem |
| Product page, no checkout | price, or trust, or an unanswered question |
| Checkout started, not finished | mechanical — payment methods, a required field, a surprise |
| Bought, then refunded | expectation mismatch; the page promised something else |
Only the second row is actually about price, and it is the one people reach for first. Repricing when the problem is in the first row changes nothing, and you will conclude the product is unwanted when nobody ever saw it.
How do you know if a video or article did anything? covers getting that picture when your analytics is on a free tier.
Should you ask people what they would pay?
Ask a different question.
Stated willingness to pay is close to worthless — people are being polite, or imagining a version of themselves with more time. What they already do about the problem is a fact: which tool they pay for, how long they spend, what they tried first.
"What do you do about this at the moment?" produces a usable answer. "Would you pay €30 for this?" produces a yes from people who will not.
What about a free version?
Only if it is a complete thing, not a crippled one.
A free item that solves one problem fully builds more trust than a sample that stops halfway. The free summary on this site is six pages of conclusions with no upsell inside it, and it exists to make the paid thing credible rather than to tease it.
The test: would you be happy if someone took the free thing and never bought anything? If not, it is not a free product, it is an advertisement, and people can tell.
What is the honest first move?
Pick a number from the three anchors, write down why, and leave it alone for a month.
The reason to write it down is that in six weeks you will want to change the price because a stranger said something, and the note tells you what evidence would actually justify that. One reasoned change beats five nervous ones.
Price what you built is a prompt for working through this properly, and The Digital Products Playbook covers the commercial side of selling files once the number exists.
Tools in this piece
| Tool | What it does | Price |
|---|---|---|
| Stripe | Payments and subscriptions. The standard, and integrating it is a matter of asking. | Free, percentage per sale |
| Shopify | Hosts the store, takes the payment and handles the tax. | From ~€29/mo |
Every tool has its own page with the price, who should skip it and what to check before paying. Some links are affiliate links.
Questions people ask
How do you price a digital product with no sales data?
Not from cost. The marginal cost of the second copy is zero, so cost-plus gives you a number with no meaning. Use three anchors instead: what the buyer would otherwise do, what the problem costs them, and what comparable things in your category sell for.
Should you start cheap and raise the price later?
Raising a price is harder than lowering one, because early buyers feel it. Starting slightly high and discounting is the easier direction, and a discount is also a reason to email your list.
Does a low price get you more sales?
Not reliably below a certain point. Under about €10 the fixed payment fee eats a noticeable share, and very low prices invite the question of why it is so cheap. Price is a signal as well as a number.
What is the fastest way to learn whether a price is wrong?
Watch where people stop. Traffic but no product page views is a message problem. Product page views but no checkout starts is a price or trust problem. Checkout starts but no payments is a mechanical problem.
Should you ask people what they would pay?
Ask what they currently do about the problem instead. What someone says they would pay is unreliable; what they already spend on an alternative is a fact.
How often should you change the price?
Rarely, and never quietly in both directions. Pick a number, leave it for long enough to see a pattern, and change it once with a reason you can state.