How to price a new service when you have no sales data yet

Pricing a new service is uncomfortable because it feels like it should be a calculation, and it is not. You have no conversion data, no churn, no sense of what buyers will tolerate. So most people do one of two things: they add a margin to their costs, or they undercut the cheapest competitor they can find. Both are reliable ways to end up at the wrong number.

The useful method is different. Your first price is not a prediction — it is a position, stated publicly, that you then test. The job is to pick a defensible position from evidence that already exists, ship it, and change it when you learn something.

Why cost-plus fails for services

Cost-plus pricing asks what the work costs you and adds a margin. For physical goods with real unit costs, that is a sane floor. For a service it usually produces a number that is disconnected from what buyers are deciding between.

The problem is that your costs are largely your own time, and your time gets faster. Price on the hours a task takes you in month one and you have built in a pay cut for every efficiency you ever find. Worse, buyers do not know or care what the work costs you. They compare your price to their alternatives and to the value of the outcome.

Cost-plus still has one job: it sets your floor. If a price does not cover your costs and leave something for the time, it is not viable no matter how attractive it looks. Calculate that floor, write it down, and then stop using cost to set the actual price.

Step one: map what buyers can actually choose

Before you can position, you need the real set of alternatives — searched for the way a buyer would search, not the way you would. For each option a buyer would plausibly find, record three things:

• The price, and how it is structured — hourly, fixed, retainer, per-unit, tiered.
• What is actually promised at that price. The scope, the turnaround, what is explicitly excluded.
• Who it appears to be for. Read the positioning language, not the feature list.

Include the alternatives that are not businesses: doing it in-house, doing nothing, and the cheap automated tool. Those are frequently the real competition, and they are the ones founders most often leave off the map.

Two things are worth noting when you cannot find a price at all. Hidden pricing usually signals either high-touch custom work or a sales process built on qualifying you first — either way, it tells you that segment is not competing on an advertised number. That can be an opening if buyers find the opacity annoying, which is the sort of thing that shows up in public complaints.

Step two: find the shape of the market, not the average

Once you have ten or so alternatives with prices, do not average them. The average of a bimodal market describes nobody. Instead, look at the distribution.

Service categories very often cluster into a cheap automated tier and an expensive full-service tier, with a thin, awkward middle. When that shape appears, the middle is not empty because nobody thought of it. It is usually empty for a structural reason — the cheap tier cannot add service without losing its margin, and the expensive tier cannot cut price without undercutting its own core business.

That structural gap is the most interesting thing on the map, because it is a position an incumbent finds genuinely hard to follow you into. If you can serve it, you are not competing on being slightly cheaper; you are the only option at that point on the curve.

If instead every option clusters within a narrow band delivering roughly the same thing, treat that as a warning. Tight price clustering with undifferentiated promises is what a competed-away market looks like, and winning it usually means winning on cost — which is a hard game for a new entrant.

Step three: pick a position and write the sentence

There are only a few coherent positions, and the failure mode is not picking one.

• Below the market, deliberately. Legitimate when your cost structure is genuinely different — automation, no overhead, a narrower scope. Not legitimate as a way to be chosen without a reason, because anyone can match a discount and the cheapest provider rarely gets to raise prices later.
• At the market, differentiated. Same price, clearly better fit for a specific buyer. This is the safest first position when you have a real specialism.
• Above the market. Requires proof — a track record, a guarantee, a speed nobody else offers, or a specialism buyers already pay a premium for. Hard on day one, but the direction you want to move.
• In the structural gap. Priced where nobody currently sits, with a scope built for that price point. Usually the strongest opening position when the gap is real.

Then write the sentence, plainly: a buyer would pay us $X rather than $Y to [alternative] because [reason]. If you cannot complete it with something concrete and checkable, you have chosen a number rather than a position, and the number will not hold up the first time someone pushes back on it.

Step four: structure the price, not just the level

How you price often matters as much as the amount. A few structural choices worth making on purpose:

• Fixed beats hourly for buyers who fear open-ended bills. A fixed price transfers risk from them to you, which is worth something — and it stops punishing you for getting faster.
• A small, well-scoped entry offer lowers the cost of trying you. When you have no reputation, the first sale is the hardest thing to get; a cheap first step is a way to buy trust, not just revenue.
• Three tiers are easier to choose from than one. A single price is a yes/no decision. Several give the buyer a "which one" decision instead, which is an easier one to make.
• Promise what you can always deliver. A turnaround or volume promise you can only hit on a good week becomes a refund on a bad one. Scope the promise to your worst realistic case, then beat it.

Step five: treat the first price as a test

Ship it, and watch what actually happens rather than what you feared would happen.

• Nobody buys, and nobody objects to the price. Usually not a pricing problem. That is a traffic, trust, or positioning problem, and cutting the price will not fix it — this is the most common misdiagnosis there is.
• People engage, then stall at the price. Now you have real pricing signal. The fix may be the level, but it is at least as often the scope, the risk, or the proof.
• Everyone buys immediately without hesitating. You are probably under-priced. Fast, frictionless yeses from every single buyer are a signal, not a victory.

Raising prices later is normal and expected — early customers bought an early-stage offer, and most understand that. Honour the old price for existing customers, announce the change plainly, and move on. That is far easier than the alternative, which is discovering a year in that you built a business on a number you picked in an afternoon and never revisited.

The short version

• Calculate your floor from cost. Then set the actual price from the market, not the cost.
• Map the real alternatives — including in-house, doing nothing, and the cheap tool.
• Look at the shape of the distribution, not the average. Structural gaps are the best openings.
• Pick one position and write the sentence that justifies it.
• Ship it as a test, and read the objections you actually get rather than the ones you imagined.