Outcomes-Based Pricing

Price your product based on measurable outcomes delivered, not usage or seats

Bret Taylor
Inside the expert network training every frontier AI model

Outcomes-Based Pricing

"The whole market is going to go towards agents. I think the whole market is going to go towards outcomes-based pricing... it's just so obviously the correct way to build and sell software." - Bret Taylor

What It Is

A pricing model where customers pay based on measurable outcomes achieved rather than usage metrics (tokens, API calls), seat counts, or subscription fees. This aligns vendor incentives directly with customer value and transforms the vendor-customer relationship from transactional to partnership.

The model becomes viable when:

  1. Agents operate autonomously (not just assisting humans)
  2. Outcomes are clearly measurable
  3. Value can be directly attributed to the product

How It Works

Traditional pricing problems:

  • Seat-based: Pay for access, not results. A tool could deliver zero value and you still pay.
  • Usage-based (tokens, API calls): A proxy for work done, but work done ≠ value delivered. As Bret notes, "tokens are similar to lines of code"—you could use a lot and produce nothing useful.
  • Subscription: Fixed cost regardless of value extracted.

Outcomes-based advantage:

  • Customer pays when value is delivered
  • Vendor is incentivized to maximize customer success
  • Easy to justify ROI in procurement
  • Transforms vendor from supplier to partner

The Sierra Example

Sierra builds AI agents for customer service. Their pricing model:

  • Resolution-based: If the AI agent solves a customer's problem completely, there's a pre-negotiated rate
  • Why it works: A typical call center call costs $10-20 (mostly labor). If AI solves it, that's measurable savings.
  • Other outcomes: Some sales agents earn commission on sales they close

Key metric: "Customers see anywhere between 50 and 90% of their customer service interactions completely automated."

How to Apply It

1. Identify the measurable outcome What specific result does your customer care about?

  • Customer service: Resolved tickets without human intervention
  • Sales: Deals closed or leads generated
  • Legal: Contracts reviewed or hours saved
  • Marketing: Qualified leads generated

2. Ensure autonomous operation The product must do the work, not just help someone do the work. Assisted productivity is hard to measure; autonomous work is clear.

3. Establish attribution Can you prove the outcome came from your product? Build measurement and tracking into the product from day one.

4. Price below the alternative If a human doing this work costs $X, price the outcome at a meaningful discount while capturing significant value.

5. Align on definitions Clear criteria for what counts as a successful outcome (e.g., "customer problem solved, customer satisfied, no escalation required").

Contrast with Consumption-Based Pricing

Consumption pricing (popular in infrastructure) charges for usage—compute hours, API calls, tokens processed. This is closer to outcomes than seats, but:

"I'm not sure a token is actually a good measure of value from AI... There's this famous story of an Apple engineer who had a bad manager who had you report how many lines of code you wrote every day... I think tokens are similar. Yeah, you used a lot of tokens, like good for you, did it produce a pull request that was good?"

Tokens measure work done, not value created. Outcomes measure value.

Requirements for Outcomes-Based Pricing

  1. Autonomous agents: Must do the job, not assist with the job
  2. Measurable outcomes: Clear definition of success
  3. Attribution: Provable connection between product and outcome
  4. Customer alignment: Customer must care about that specific outcome

Why the Market Will Move This Way

Bret predicts outcomes-based pricing will become dominant because:

  1. CFO clarity: Easy to justify and measure ROI vs. opaque SaaS bills
  2. True partnership: Aligns vendor and customer interests
  3. Market pull: Companies that offer it will win against those that don't
  4. Productivity reality: Agents actually do work (vs. software that helps humans work)

"The whole market is going to go towards outcomes-based pricing, not because it's the only way, but the market is going to pull everyone there because it's just so obviously the correct way to build and sell software."

Source

  • Guest: Bret Taylor
  • Episode: "Inside the expert network training every frontier AI model"
  • Key Discussion: (01:04:45) - Detailed explanation of Sierra's outcomes-based model
  • YouTube: Watch on YouTube

Related Frameworks

Back to Execution