Insights / How to Pivot to Outcome-Based Pricing (It's a GTM Change, No…
How to Pivot to Outcome-Based Pricing (It's a GTM Change, Not a Billing One)
Intercom charges 99 cents per resolved ticket and built a nine-figure business on it. The pricing is the easy part to copy.
Alice B
Outcome-based pricing is charging for a result your product delivers rather than access to your product: per resolved ticket, per booked meeting, per qualified lead, instead of per seat per month. Intercom's Fin agent is the cleanest proof it works at scale, charging 99 cents per resolved conversation and growing from $1M to over $100M in annual recurring revenue on that single number.
Switching to outcome-based pricing changes your sales motion, your value proposition, your onboarding, and your customer success function. It's a go-to-market transformation wearing a billing change's clothes.
Every invoice under this model is a proof-of-value statement. That's the whole appeal, and it's why the best example is worth studying before you touch your own pricing page. Intercom's Fin charges 99 cents per resolved conversation, now handles more than 80% of support volume for the customers running it, resolves around a million issues a week, and grew from $1M to north of $100M in ARR on that model, backed by a performance guarantee of up to $1M if resolution targets aren't met. No resolution, no charge. The product investment and the pricing structure became the same thing: the better Fin gets, the more it bills, automatically.
Intercom's Fin charges $0.99 per resolved conversation and grew from $1M to over $100M in ARR on that model, resolving roughly 1M issues a week.
The clearest evidence that outcome-based pricing scales, and the easiest part of the story to copy.
Source: Intercom; Gleap analysis of Fin pricing, 2026.
Why now, and not five years ago
Because agentic AI breaks the three assumptions that made seat-based pricing make sense. Seat-based pricing assumed software was passive, that it gave you information and a human did the work; agents act instead. It assumed value was diffuse across a team and couldn't be pinned to the tool; a resolved ticket is binary and obviously the agent's doing. And it assumed you couldn't attribute outcomes cleanly; now you can. Bessemer put the principle bluntly: products should get paid for outcomes, not access. Gartner's projection is that by 2030, at least 40% of enterprise SaaS spend will shift toward usage, agent, or outcome-based models, up from under 10% today. The transition is early, which is exactly why getting it right now is worth something.
There's a margin reason too, and it's not optional. AI-native products tend to run gross margins of 50% to 60%, against 80% to 90% for traditional SaaS, because inference costs money on every call. You need a pricing model that expands with the value you deliver, or the economics simply don't close. Outcome-based pricing is partly a philosophy and partly a survival mechanism for AI margins.
Pressure-test your pricing as one of twenty-two commercial levers
Eight questions. Five minutes. No email required.
Take the GTM self-assessmentThe five-step pivot
Here's how to actually make the move, in five steps. The order matters; most of the pain comes from doing step five after step one instead of alongside it.
Define what an "outcome" actually is
This is where most pivots quietly fail. A good outcome passes three tests: it's binary (it happened or it didn't), observable (you can log it without the customer self-reporting), and caused (you can defensibly argue your product caused it, not the market or the sales team). "Customer is satisfied" fails all three. "Resolved ticket" passes all three, which is why Intercom could build on it. Run every candidate outcome through the Outcome Test before you price a cent against it.
Run the unit economics before you announce anything
Outcome-based pricing makes your cost of goods variable and tied directly to output volume. Model four numbers: average inference cost per attempt including the failed ones you eat, resolution rate now versus projected, margin per outcome at target volume, and what your ARR looks like with a plus-or-minus 10% swing in that resolution rate. Find that out in a spreadsheet, not in a board meeting.
Don't kill the old model. Build a ramp
Salesforce is the template: Agentforce reached around $800M in ARR running three pricing models at once, per-conversation at roughly $2, a credit system per action, and traditional per-seat for teams that want predictability. CFOs hate variable invoices even when they're cheaper on average, so the model most companies land on first is a hybrid: a base subscription for predictability plus outcome overage for upside.
Instrument before you price
You cannot charge for an outcome you can't measure, and you can't dispute an invoice you can't evidence. Build outcome tracking into the product before the pricing change, not after. Establish baselines with current customers so you know their resolution rate today, and create the audit trail customers will ask for the first time they query a bill.
Handle revenue recognition early
Outcome-based payments are variable consideration under ASC 606, which means you can only recognise revenue when it's highly probable a significant reversal won't occur. Get your accountant in the room before you go to market, not after the first quarter closes.
The methodology: The Outcome Test
A candidate outcome is priceable only if it passes three tests: binary (it happened or it didn't), observable (you can log it without customer self-reporting), and caused (you can defensibly attribute it to your product, not the market or the sales team). "Resolved ticket" passes all three. "Customer is satisfied" fails all three.
The hard truths the case studies skip
Around 45% of early adopters hit initial margin pressure during the transition, per Deloitte. About 70% of companies struggle to accurately measure outcomes in their contracts, per McKinsey, usually because the instrumentation in step four was an afterthought. And outcome disputes are real in a way seat-based pricing never produced: "we resolved it," "we disagree," is a new category of customer friction you're signing up to manage.
Three ways to charge for outcomes
| Approach | Who charges this way | Where the risk sits |
|---|---|---|
| Per resolution (shared risk) | Intercom Fin, $0.99/resolution | Split: customer pays only on success, vendor eats failed attempts' compute |
| Per resolution, full vendor risk | Zendesk, charges only on full resolution | Entirely on the vendor; a strong enterprise sales argument and incentive to ship AI that works |
| Multi-model ramp | Salesforce Agentforce, ~$2/conversation + credits + seats | Spread across models; customer picks predictability or upside |
The thing to hold onto is that none of this is really about the number on the invoice. It's a GTM transformation, and that's the underrated part. Pricing is one of twenty-two levers in the commercial layer, and when you move this one, it pulls on the others: your sales motion has to sell an outcome instead of a feature, your onboarding has to drive the customer to the outcome fast because that's when you get paid, and your customer success team becomes the function that protects revenue rather than just renewals.
The companies that treat this as a billing tweak will ship a new pricing page and wonder why nothing else got easier. The ones that treat it as a go-to-market rebuild will find that every invoice starts doing the selling for them.
Frequently asked questions
What is outcome-based pricing?
Outcome-based pricing charges for a result your product delivers, such as a resolved support ticket or a booked meeting, rather than per seat or per month. Intercom's Fin is the best-known example at $0.99 per resolved conversation.
Is outcome-based pricing just usage-based pricing?
No. Usage-based pricing charges for activity (API calls, actions, seats used). Outcome-based pricing charges only when a defined result is delivered. A good outcome is binary, observable, and causally attributable to your product, which usage metrics usually aren't.
How do I switch my SaaS to outcome-based pricing?
Five steps: define the outcome (binary, observable, caused), model the unit economics before announcing, build a ramp rather than killing the old plan, instrument the product to measure and evidence outcomes, and handle ASC 606 revenue recognition early with your accountant.
Why is outcome-based pricing happening now?
Because agentic AI makes outcomes attributable in a way passive software never did, and because AI-native margins (50-60%) need pricing that expands with value. Gartner projects 40% of enterprise SaaS spend shifts to usage, agent, or outcome-based models by 2030, from under 10% today.
What are the risks of outcome-based pricing?
Variable cost of goods tied to output, initial margin pressure (around 45% of early adopters, per Deloitte), difficulty measuring outcomes (about 70% of companies struggle, per McKinsey), outcome disputes with customers, and ARR-predictability questions from investors.
What counts as an "outcome" I can charge for?
One that passes three tests: binary (it happened or it didn't), observable (you can log it without customer self-reporting), and caused (you can defensibly attribute it to your product). A resolved ticket passes; "customer satisfaction" fails.
