Pricing
Startup pricing: five steps to your first price
Some sections describe the law in Germany (as of July 2026). General information, not legal advice. Other countries regulate this differently.
A price is built in five steps. First: work out your variable cost per unit and your fixed costs. Second: collect the prices of five to eight providers your customers compare you against. Third: put a money figure on what the customer gains. Fourth: form a range between that floor and that ceiling. Fifth: test a price in the upper third of it. Cost tells you the floor, never the price. Pricing is also the strongest profit lever there is: on the S&P 1500 average, a one percent higher price lifts operating profit by around eight percent, roughly three times the effect of one percent more volume.
Why the price is the strongest lever
Three ways to earn more: charge more, spend less per unit, sell more. They are not equally powerful. The price change goes into profit almost untouched, while extra volume brings extra cost with it.
Effect on operating profit
Figures: McKinsey, "The Power of Pricing", average of the S&P 1500. The model assumes volume stays constant, which in practice it rarely does entirely. The order of the three levers holds up regardless.
The three methods, and what each is good for
You need all three, but not as alternatives: cost sets the floor, value sets the ceiling, and the competition tells you where inside that space you are being read.
The floor in figures: understanding break-even, with an interactive calculator →
The five steps
Half a day of work in total. Step 3 is the one most founders skip, and it is the one that decides whether you price on cost or on value.
Calculate the floor
90 minutesList every variable cost per unit sold: materials, shipping, payment fees, commissions, and for software also hosting and support per customer. Add your monthly fixed costs. The floor is the price at which a sale starts contributing anything at all.
Result: a number you never go below. It is not your price, it is your floor.
Collect reference prices
60 minutesGather the list prices of five to eight providers your customers actually consider. Always note the scope alongside the price, otherwise you are comparing numbers without content. If a provider has no pricing page, the price is usually negotiable and high.
Result: the range the market moves in, and a decision about which end of it you want to sit at.
Put a number on the benefit
2 hours, spread across conversationsAsk five potential customers what the problem costs them today: in time, in lost revenue, in workarounds they already pay for. Convert those answers into an annual figure. Someone spending eight hours a month on a task your offer cuts to one saves roughly 84 hours a year. That figure is your ceiling.
Result: a sentence of the form "the customer saves X per year". Without it, cost is the only thing you can price on.
Set the range, pick the model
60 minutesYour range sits between the floor and the value figure. Choose a point inside it and a pricing model (see below). Rule of thumb for founders: aim at the upper third of the range. Going down later is always open to you, going up costs you existing customers.
Result: a concrete price plus a justification you can say out loud in a sales conversation.
Test it, do not poll it
2 to 4 weeksSay the price in real sales conversations and watch the reaction. If nobody buys, the price is not automatically at fault: usually the benefit was never made explicit. If everybody agrees immediately, the price was too low. A healthy rate is one where some enquiries do fail on price.
Result: an evidenced decision instead of a gut number, and a date for the next price review.
Four pricing models
The model decides how the price is perceived, not just how much comes in. Pick the one that matches how the benefit arises.
One-off price
One price per delivery. Easy to understand and to invoice, but every unit of revenue has to be won again.
Tiers
Three levels with clearly different scope. The middle one gets picked most often because it wins the comparison. Two levels is too few, five confuses.
Subscription
A recurring amount per month or year. Predictable, but it requires the benefit to recur too, not just happen once.
Usage-based
Price per volume, per user or per transaction. Grows with the customer, but makes their bill unpredictable.
The price test with four questions
Asking "what would you pay?" produces polite, useless answers. This set of four questions, known as the Van Westendorp price sensitivity meter, brackets the range from four directions instead. Ask them in this order, without showing a price first.
- 1At what price would the offer be so expensive that you would not buy it?
- 2At what price would it be expensive, but still worth thinking about?
- 3At what price would it be a bargain?
- 4At what price would it be so cheap that you would doubt the quality?
Answers 1 and 4 mark the outer limits, 2 and 3 the range people actually accept. Statistically the method needs a few hundred responses. With 20 to 30 conversations you get a direction, not a proof, and that is still far better than guessing.
Where the reference prices come from: competitor analysis in 5 steps, on no budget →
Five mistakes that cost margin
- Costing only. Cost accounting gives you the floor, never the price. Pricing that way gives away exactly the margin the customer would have paid.
- Starting low to win customers. An entry price trains the customers it attracts. Raising it by 30 percent later costs more conversations than a higher price would have from the start.
- Discounting instead of offering a smaller package. Every discount granted is a lasting statement about value. A cheaper tier with less in it is the better answer.
- Never touching the price again. Costs rise every year, prices do not rise by themselves. One fixed date a year is enough.
- Hiding the price. With no price shown, part of your prospects leave before you ever speak to them. Visible prices pre-qualify, and that saves you time.
Frequently asked questions
How do I calculate my hourly rate as a service provider?
Work backwards from your target: desired annual income plus operating costs plus a reserve for tax and social contributions, divided by the hours you can actually bill. The mistake is nearly always in the denominator: out of roughly 1,700 working hours a year, solo operators can typically bill about half, with the rest going into acquisition, proposals, bookkeeping and training.
Should I start with a low introductory price?
Only for a limited time and named as such, for example a price for the first ten customers in exchange for feedback. A permanently low entry price attracts price-sensitive customers who leave at the first increase, and gives you a distorted picture of what your actual market is willing to pay.
How often should I review my prices?
Once a year on a fixed date, plus whenever costs rise noticeably or the scope of what you deliver changes. A review needs three numbers: the share of enquiries that fail on price, the contribution margin per sale, and where competitors sit.
What are the legal rules on displaying prices?
In the EU and the UK, prices shown to consumers must be the total price including VAT; net-only prices to consumers are not permitted. In the US, sales tax is added at checkout and varies by state and city. Registration thresholds for VAT and sales tax exist in every jurisdiction and change regularly, so check the current figure for yours before you set a price. This is not tax advice.
Do prices ending in 9 work?
In price-sensitive consumer markets they measurably do, because they read as a deal price. In premium and business settings that is precisely the drawback: 4,999 signals a discount where reliability is what is being bought. Round numbers fit better there.
Read next
Have the numbers checked before you commit
The ANALYST agent works through unit economics, margin and break-even for your specific idea, while the VALUE ANALYST rates how strong your promise really is. From € 5.99, in minutes.
Assess my idea →