You're Charging Too Little

Most software is underpriced, and the companies that make it rarely find out. Underpricing doesn’t announce itself — there’s no error message, no angry customer, no line on the P&L labelled “money we could have had.” The product sells, revenue grows, and the gap between price and value quietly compounds.

This field guide from Fathom opens with a survey of 1,200 software companies in early 2026, and one number that frames everything after it: 92% set their prices without any structured research into willingness to pay. They copied a competitor, added a margin to their costs, or picked something that felt right.

Price is the highest-leverage number you have

A 1% improvement in price drives more profit than a 1% improvement in volume, cost, or acquisition. The reason is mechanical — a price increase falls almost entirely to the bottom line, because you have already paid to build and sell the product. There is no additional cost to serve a customer who pays 10% more.

And yet pricing gets a fraction of the attention acquisition does. Companies spend six months optimising a signup funnel to lift conversion two points, while the price at the end of that funnel — the number that determines the value of every single conversion — was set in an afternoon three years ago.

It also leaks into everything else. Price too low and you attract bargain-hunters who churn hard and file the most support tickets; the cheapest customers are almost always the most expensive to serve. Customers paying a price aligned with the value they get stay longer, because the relationship makes sense to them.

Two of the three ways to set a price are traps

Cost-plus is the accountant’s trap. Your costs have nothing to do with your customer’s value. A tool that costs €2 per user per month to run might save a customer €4,000 a month; cost-plus would have you charge €5 and feel clever about the margin. Software’s marginal costs are near zero, which anchors you to the one number that matters least.

Competitor-based is the sheep’s trap. Remember that 92% — when you copy a competitor you are not inheriting insight, you are inheriting their guess and compounding it across the category. Competitor pricing is a useful input for knowing where the market’s mental anchors sit. It is a catastrophic foundation.

Value-based is harder, and it is the only method that can be right rather than merely defensible.

Decide what you charge for, before how much

The value metric is the unit your pricing scales on: per seat, per contact, per gigabyte, per invoice sent. A good one aligns with the value the customer gets, is easy enough to understand that they can predict their bill, and grows as the account grows — which is the engine behind net revenue retention above 100%.

Per-seat is the B2B default and quietly wrong for many products, because it charges for access rather than value. A billing tool’s value scales with invoices processed, not with the two finance people who log in. Price per seat on a product whose value scales differently and two bad things happen: customers ration seats to control cost, limiting how far your product spreads inside the account, and your revenue decouples from the value you deliver.

The question to ask is simple: when does this customer succeed, and what number goes up when they do?

Then actually ask what they’d pay

The step almost everyone skips. The Van Westendorp Price Sensitivity Meter — four questions, a survey technique from 1976 — takes a few hundred responses and an afternoon of analysis to produce a defensible price range. Most companies have never run it once.

The insight underneath it is that willingness to pay is never one number. A solo freelancer and a 500-person enterprise don’t get the same value, don’t have the same budget, and shouldn’t pay the same price. Charge them identically and you scare off the freelancer while massively undercharging the enterprise — usually both at once. The answer isn’t one perfect price; it’s segmentation, using features and usage tiers as fences that let each segment self-select.

The report closes with a ten-question pricing audit and a scoring guide, which is the fastest way to find out whether you’re in the 92%.

About this example

This is a real report from a pricing consultancy, published as an ungated PDF — and it is in this library because it shows what a finished Unplain document looks like when the content is dense: comparison tables, a do/don’t grid, chapter openers, and a scored audit, none of which were laid out by hand. The manuscript was written as a plain document and Unplain produced the rest.