Value-based Pricing

The oldest way to set a price is to add up what it costs and put a margin on top. For physical goods that works. Making one more unit consumes more materials and more labor, so cost is both the floor under the price and a sense of scale.

For software the arithmetic collapses, because one more user costs almost nothing extra. Pagedeck started out pricing on cost. Servers and storage came to about $1.50 per account per month, so a 30 percent margin put the price at $2.

The number was tidy and the reaction was strange. Sofia heard the same thing three times from enterprise prospects: at that price, is something missing? One of those companies had two people assigned to collecting and tidying documents, spending over $2,300 a month on it. To that company Pagedeck was not a two-dollar item.

Cost is a fact about the seller's situation. What a customer can pay depends on what the product replaces on their side. Value-based pricing accepts that the two numbers have nothing to do with each other and moves the basis to the second one.

Convert what the customer gains into money first, then ask for a share of that amount. There are four steps.

1. Find what changes. There are only four kinds: time goes down, headcount goes down, something that was being lost stops being lost, or revenue goes up. Pagedeck lands on the first two — the work of gathering and tidying documents shrinks.
2. Convert it to money. This is where numbers are needed. If each person saves 12 hours a month and that time costs $23 an hour, each person generates about $280 a month. That figure has to come from the customer; estimated in-house it comes out inflated. Asking for it is also the fastest way to learn whether the customer believes in the saving at all.
3. Decide what share to take. The common band is 10 to 30 percent of what the customer gains. Ten percent of $280 is about $28 a month. Take all of it and the customer has no reason to switch; take far too little and the difference is money left on the table.
4. Split by segment. The same product is worth different amounts to different people. What one person working alone saves and what a team of twenty saves differ by an order of magnitude, which is why there cannot be a single plan.

The billing model, the structure by which money is collected, falls out of this too. The unit should grow in the same direction as the value the customer receives. Pagedeck counts seats, because a bigger team means more documents to gather and more hours saved. Counting storage attaches the price to a number unrelated to the value, and the first customer who notices asks why they pay more for the same benefit.

None of this only pushes prices up. Cost is the floor and what the customer gains is the ceiling, so the range in between is where a price can sit. Sometimes the ceiling comes out below the cost. The answer then is not to raise the price but to drop that segment from the list of people you sell to.

The first place is building the plan table, and specifically the upper rows, which do not stand up without this calculation. Free and entry-level paid rows can be set against competitors, but the enterprise row has nothing comparable, so the only route is coming down from the amount the customer gains.

The second is enterprise negotiation. Sofia asks for the customer's numbers before saying a price: how many people are on this work now, how many hours a month it takes. The same figure reads differently after those answers arrive, because it becomes $300 against the $2,300 a month they were already spending.

The third is raising prices. Justify an increase with cost and the customer hears a demand for more money over something that has nothing to do with them. Show how many hours were saved over the past year and adjust against that, and the conversation starts somewhere else.

The fourth is product decisions. Features that increase the hours a customer saves separate from features that don't, and the second group can be nice to have without being grounds to raise a price, so the calculation quietly becomes a filter on the roadmap.

There are poor fits. For a product an individual uses, saved time is hard to convert into money, and for something used for fun or taste there is nothing to convert at all. In those cases setting a band from comparable products is more practical, or finding a range with a Van Westendorp survey that asks where a price starts feeling expensive and where it starts feeling cheap.

"Let's set this plan on value. There's no comparable row at any competitor."

With nothing to compare against, the price comes down from the amount the customer gains.

"We can't quote until we know how much this saves them."

Without a figure to convert, the method has nothing to work with.

"Charging by seat is right. Charging by storage doesn't track the value."

Align the billing unit with whatever grows as the customer's gain grows.

At Pagedeck, Grace and Sofia run this calculation together. Sofia collects the hours and headcount currently spent on the work from five or six customers, and Grace decides what share to charge. Mia carries those numbers back to the product. Measuring saved hours means the activity has to be recorded somewhere, so claiming the value requires a screen that shows it first.

The most common reading is that this is the expensive option. Prices often do land above cost, so it looks that way. What the method fixes is the basis, not the level. Measuring the ceiling and finding it lower than planned, then cutting the intended price, is the same method producing a result. Treat charging more as the goal and you end up raising numbers without the calculation, which leaves a price nobody can explain.

The second is thinking you can simply ask customers what they would pay. The value accrues to them, so they ought to know best. But people cannot name a fair price. They tend to say something slightly below what they pay now, or whatever number seems polite. What to ask for is the raw material, not the price: how many hours this takes today, how many people are on it. Those answers come back consistent, which is the whole reason the question is worth asking, and the price is something you calculate from them.