The one confusion. Outcome pricing isn't charging more for the same work. It's selling a result instead of the inputs that make it. Hours and tokens sell effort; outcomes sell labor done. And the hard part isn't the price, it's the measurement: whoever owns a baseline both sides trust owns the deal.
The moves that matter
| Move | Think of it as | What actually happens |
|---|---|---|
| Name the outcome | The finish line | One result the buyer already cares about and can count: hours saved, revenue added, errors cut. If they can't count it, it isn't an outcome yet. |
| Sign the baseline | The "before" photo | Write down today's numbers and how they'll be measured, and both sides sign it before any work starts. Every result is checked against this page. |
| Share the gain | A royalty | You're paid an agreed share of the measured result, on a set schedule, with a cap, a floor and an end date. |
| Keep a base fee | Salary plus bonus | A modest fixed fee or retainer pays for the work itself; the gain share sits on top. Results come in lumpy, and rent doesn't. |
| Offer the picker | Choosing how you pay for ads | Let the buyer choose hourly, fixed fee or outcome. Choice removes "you're gambling with my business", and what they pick tells you what they value. |
Why the market is moving this way
Token sellers sell compute; outcome sellers sell labor. Ad pricing climbed the same ladder, from paying per view to per click to per sale, and AI pricing is following it.
- Big vendors are reported to be moving. One AI platform is said to price its agents on the revenue they generate; another reportedly lets its largest customers pay only when a job completes.
- Responsibility sells at a different price. A CRM that struggled at $50 a year reportedly sold at $20–30K once wrapped in "total success". Buyers pay far more for a result than for a tool.
- It reveals what work is worth. When buyers pick outcomes, they show which jobs are worth $10K and which $1M. Expect a pricing picker, like ad platforms: pay by compute, by tokens or by results.
- It can keep people employed. A small bank spending a few dollars on tokens is nobody's priority. Priced on results, the vendor cares whether the bank actually grows, and the bank's people stay.
From a Moonshots podcast segment (Alex Wissner-Gross). Treat the vendor and price claims as reported, not verified.
Safe to try
None of these can lose you money; they test the idea before you bet on it.
| Safe move | Why it cannot hurt |
|---|---|
| Put a picker on your pricing page | Hourly and fixed stay available; outcome is just one more choice. |
| Ask a client what result they'd pay for | A conversation. You learn what they count, and whether they count it at all. |
| Price the first deal mostly fixed | Outcome money arrives late. Bank cash first, add a share once there's trust. |
| Write the baseline before any number is promised | Nothing is owed or claimed until both sides have signed what "before" was. |
The everyday loop
Agree the outcome, sign the baseline, do the work, measure on schedule, get paid your share, and decide together whether to renew. The measurement, not the price, is what you get better at each time.
Practice, start to finish
- A client wants their support inbox answered faster. You agree the outcome: median reply time. One number both of you can pull from their own tool.
- Sign the baseline: today it's 20 hours, measured from the help-desk report. A one-page agreement with the number, the source and the dates.
- Set the price: a fixed fee for the build, plus a monthly bonus for each hour cut, capped, with a floor and a 12-month end date. They know the most they'll pay; you know the least you'll get.
- Build and launch. Replies speed up, and the client sees it in their own report, not yours.
- Each month, pull the number, send a short statement, invoice the share. No debate: same report, same method as the baseline.
- At month 12, renew, re-baseline or part as friends. No open-ended claim on their business.
When you make a mess
| Situation | Way out | What it does |
|---|---|---|
| The client disputes the result | Go back to the signed baseline and the agreed source | Turns an argument into arithmetic. |
| Something else moved the number | Agree up front what counts as "outside" (a price rise, a new hire) | You're paid for what you changed, not luck either way. |
| Gains fade after the easy wins | The floor protects you; the end date closes it cleanly | No one is stuck paying for gains that faded. |
| Too small to measure | Price it fixed or hourly instead | Not every job needs outcome pricing. |
Leave this alone: promising an outcome before there's a signed baseline. Without one, every payment becomes a negotiation, and the client decides what you earned.
Why the way out works. Everything rests on the baseline both sides signed before the work began. With it, outcome pricing is a measurable deal. Without it, it's a hope.
Words you will see
Outcome-based pricing
Paid for a result delivered, not for time or materials used.
Gain share
Your agreed percentage of the measured result: savings or new revenue.
Baseline
The signed "before" numbers that every result is measured against.
Cap / floor / time-box
The most you can earn, the least you'll accept, and when it ends.
Time and materials
Paying for hours and costs as they happen; the opposite end from outcomes.
CPM / CPC / CPA
Ad pricing per view, per click, per sale; a map for compute, tokens and outcomes.
Deliberately not here
Contract language, insurance and tax treatment (get a professional before your first outcome contract), measurement methods for specific industries (they come from doing a few), and enterprise-scale deals (the moves are the same; the baselines get political). Start small, measure honestly, and the rest can wait.
Outcome-based pricing · two-page primer · October 13, 2026 · Rich Price