How to price a new product when you have no data

Pricing a product nobody has bought yet is a decision made without the one input that would settle it: evidence of what people will pay. So the price gets set by whatever is nearest to hand — the cost plus a comfortable-sounding markup, or a competitor's number, or a figure that simply feels reasonable. Any of those can land somewhere workable by accident. The problem is that the first price is much harder to change in one direction than the other.

Raising a price on people who already bought at the old one requires explanation and costs goodwill, and some of them leave. Lowering a price is easy, welcomed, and immediately increases volume. That asymmetry is the single most useful thing to know when you have no data, because it tells you which way to err. Starting above where you expect to land leaves room to discount, run an introductory offer, or quietly settle lower. Starting below leaves you with customers anchored to a number that does not work and no graceful way out.

Work out what one unit genuinely costs you — the goods, the packaging, the delivery, the payment charge, the commission — and you have the level below which every sale is a donation. Add a share of your fixed costs at a realistic volume and you have the level below which the business does not survive even at good volume. That is the floor, and it is genuinely useful: it rules out a whole region of prices with certainty, which is more than most pricing methods do.

What it cannot do is tell you the price. Cost-plus pricing sets the number by reference to your own efficiency, and your customer has no interest in that. A supplier who gets you a better rate does not make your product less valuable to a buyer, and an expensive process does not make it more valuable. Businesses that price only from cost end up leaving money behind on products customers value highly and stubbornly overpricing ones customers do not, because the input they are reading has nothing to do with demand.

What comparable products sell for tells you what the market has trained buyers to expect, which is real information even though it is not a target. Gather it properly rather than casually: note the actual transacting price after routine discounts, what is bundled in — delivery, installation, warranty, support, returns — and which customer the competitor is serving. A number without those attachments is not comparable to yours.

The caveat is that you cannot see their cost base, and their price may be sustained by something you do not have: scale, an owner-financed loss, a different sourcing arrangement, or a second product the first one exists to sell. Matching a price set by a structure you do not share is how businesses end up working hard at a loss. Read competitor prices as a description of buyer expectations, and treat a large gap between their price and your floor as a question to investigate rather than a verdict to accept.

The upper limit on a price is what the product is worth to the buyer, which for business buyers is often calculable. If something saves four hours a week for someone whose time has a known cost, or prevents a category of loss with a known frequency, the value is arithmetic and the buyer can follow it. For consumer purchases the value is less tractable but no less real: convenience, appearance, reliability, avoided hassle, the way a purchase makes someone feel in front of other people.

The practical move is to write down, in one sentence, what the buyer gets that they did not have before, and then ask what the alternatives to getting it cost — including doing nothing, doing it themselves, and hiring someone. Those alternatives are the buyer's real comparison set, and they are frequently more expensive than the competitor product you were benchmarking against. Value-based pricing is not a licence to charge more; it is the discipline of finding the ceiling so you know how much of the range between floor and ceiling you are choosing to leave on the table.

Asking "would you buy this at ₹2,000?" produces answers close to worthless, because agreeing is free and pleasant and costs the respondent nothing. Better questions make the respondent give something up. Ask what they currently spend solving this problem and how — a real figure from the past rather than an intention about the future. Ask what they last paid for something in this category, and what made them choose it. Ask at what price it would be expensive enough to think twice, and at what price they would doubt the quality; the gap between those two answers describes a usable range and is far more stable than a yes-or-no.

Stronger still is a test that involves a commitment: a pre-order at a stated price, a deposit, a waitlist that asks for a card, a landing page with the price visible and an actual checkout button. Even a handful of responses to a price-visible page tells you more than dozens of survey agreements, because clicking through when a number is on screen is a behaviour rather than an opinion. And if you have enough traffic to split it, showing two prices to two comparable groups measures the trade-off between price and conversion directly.

A reasonable way to open is to set a placeholder inside the range the three methods leave you: above the floor, within sight of buyer expectations, well below the ceiling if the ceiling is far away. Say to yourself in writing that this is a placeholder, what evidence would move it, and when you will look again. Written down, it stops being the accidental permanent price it otherwise becomes, and it stops the first customer conversation from settling the question by default.

Be honest about what the placeholder does not tell you. Early buyers are not typical buyers; they are the people most eager for the thing to exist, and their willingness to pay runs ahead of the market's. Low volume cannot distinguish a pricing problem from a distribution problem, an awareness problem, or a product that is not ready. And no pricing method available before launch can tell you the price customers will actually pay, only bound the region where it plausibly sits. The bounds are worth having; treating them as precision is not.