How to test your price while the product is still rough

Price usually gets set at the last minute: look at competitors, pick something in the middle, knock a bit off so as not to scare anyone. Then nobody touches it for years. Yet price is one of the few things you can test before the product is ready, and one of the few that change your economics immediately.

Money in interviews

Asking “how much would you pay?” is pointless. People name a number that feels polite and then don’t pay it. Ask about money that’s already being spent.

• How much do you currently spend on this task, on tools, contractors, staff time?
• Who in the company decides what to pay for, and out of which budget?
• What did you last pay for that was similar, and how was the decision made?

The answers don’t give you a price. They give you a ceiling and a frame. If someone spends ten hours of their own time a month on the problem, you know what it costs them. If they spend nothing and don’t plan to, no price will be right.

At the end you can name a price, as a question rather than an offer: “If this cost such-and-such, what would you think?” Watch the reaction, not the words. A pause, a question about terms, an attempt to haggle are good signs. An instant “sounds fine” often means you named too little.

Anchors

Nobody judges a price on its own; they judge it next to something. You can use that honestly.

The alternative as the anchor. If the product replaces half a day of a bookkeeper’s work each month, show that comparison. A subscription next to a salary looks different from a subscription next to other subscriptions.

A top tier. An expensive plan for large teams makes the middle one look reasonable, even if hardly anyone buys the top one. But it has to be real, with things large teams actually need, not decoration.

Annual billing. A discount for paying a year up front shows how many people are confident enough in the product to pay in advance. That’s a signal in itself.

The fake door

The cheapest pricing test is a pricing page before there’s anything to sell. Someone picks a plan, clicks “pay”, and sees an honest message: payments aren’t open yet, leave your email, we’ll write to you first and hold this price for you. You learn how many people reached the button and which plan they chose.

A few rules to keep it from becoming a trick. Don’t take money for something that doesn’t exist. Don’t pretend checkout is broken: say plainly that the product isn’t ready. And keep the promise: people who left their email get that price.

You can show different prices to different visitors, but carefully: if two friends see different numbers on the same page, you’ll be the one explaining. It’s safer to vary price over time, one price this week and another the next, or by segment on separate pages.

Raising prices for new users

If the product is already selling, the simplest move is to raise the price for new users only. Existing customers keep their terms, which is fair and removes the biggest fear. Meanwhile you watch what happens to paid conversion at the new price.

Look beyond conversion. Revenue per visitor can go up even if fewer people pay. Retention at a higher price is sometimes better, because an expensive product gets bought by the people who really need it. And it matters who dropped off. If the people you lost were never your segment anyway, that’s not a loss.

Signs your price is too low: almost nobody negotiates, nobody complains, and purchases happen without anyone needing sign-off. A few complaints about price are a normal part of a healthy product.

What follows

Price is a hypothesis like everything else in an early product. You can test it, change it and test it again, and it’s better to do that early, while users are few and every change is visible.

SUBXCAT can help plan these tests and build a pricing page that runs them without touching the core product.