Value-Anchored Outcome Pricing
Price the outcome, not the seat; set the number from customer value and let cost be your problem.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 90%
For AI products, McCabe priced Finn per resolved outcome (99 cents per resolved ticket) rather than per seat, aligning revenue 100% with the value delivered. The price is anchored to customer-perceived value — not cost — and set at the point that maximizes both fairness and take, betting that costs will fall over time. It doubles as a brutal viability test: if a customer won't pay the outcome price for excellent work, there's no business.
Origin
Eoghan McCabe on pricing Intercom's Finn AI agent; validated on the podcast against pricing expert Madhavan Ramanujam's principles of simpler, outcome-based pricing.
Core principles
- 01Align revenue 100% with the value the product attains (a resolved outcome)
- 02Set price from value, not cost — 'cost is our problem'
- 03People value digital work less than human work even when it's better; price into that reality
- 04Bet that unit costs will fall, so a value price that's thin on margin today widens later
- 05The willingness-to-pay threshold is also a go/no-go test for the business
How to run it
- 1
Define the unit of value
Pick the outcome the customer actually cares about and can measure — for Finn, a resolved customer ticket — and make that the billable unit so revenue tracks delivered value.
Pro tip Choose the metric competitors are graded on anyway (resolution rate) so your pricing and your quality obsession point the same way.
- 2
Research incumbent cost-to-serve
Find what customers currently spend to achieve that outcome by other means — Intercom found SaaS businesses spent $20-30 per ticket resolved (they spent ~$22), consumer businesses ~$5.
- 3
Anchor the price to perceived value, discounting for digital
Set the price below the human cost to be obviously fair, but account for the fact that buyers discount digital labor versus human labor even when it's better — Intercom considered $25, $10, $5 before landing at 99 cents.
Pro tip Find the 'nexus between us earning the most and it being the most palatable' rather than either extreme.
Watch out Don't price from your own cost — treat falling model costs as your margin upside, not the customer's anchor.
- 4
Use the price as a viability test
Treat the threshold as a filter: if a customer won't pay the outcome price for work you deliver perfectly and instantly, conclude there is no business there.
In the wild
Against a backdrop of Intercom's historically hated, complex pricing, McCabe priced Finn at 99 cents per resolved ticket — anchored to the ~$22 businesses spent per human resolution and the belief digital work is discounted — while accepting thin early margins on the bet costs would drop.
→ Costs fell substantially, margins became 'more than worth our while,' and Finn is rated #1 in G2 with the highest resolution rate, biggest by customer count and revenue in its category.
Common mistakes
Pricing from cost instead of value
McCabe insists price should come from value, not cost — 'cost is our problem' — because early AI unit costs are high and falling, so cost-based pricing would misprice the actual value delivered.
Assuming digital work is valued like human work
Buyers discount digital labor even when it's better and more consistent; pricing as if customers value it equally overshoots willingness-to-pay.
Is it for you?
Best for
Founders pricing an AI agent or automation product that delivers a discrete, measurable outcome and want revenue aligned with value.
Not ideal for
Products where the unit of value is fuzzy or unmeasurable, or where costs won't fall enough to make a thin value-anchored margin sustainable.
From the transcript
“we charge 99 cent to resolve tickets”
“I always believe that value should that pricing should come from value and not from cost. Cost is our problem.”
“we found that many SAS businesses were spending between 20 and $30 per ticket resolve. We were spending 22.”
“we started to sense that people just wouldn't value the digital work as much as the human work”
“if someone is not prepared to pay 99 cent for us to rapidly and elegantly, perfectly and excellently solve their customer's problem. We need to…”
From the episode
How Intercom rose from the ashes by betting everything on AI
Eoghan McCabe (founder and CEO)