Frameworks, core principles and top case studies for SaaS pricing, learnt and refined over 28+ years of SaaS-monetization experience.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
.png)
Eoghan McCabe walked back into Intercom in October 2022 and found a company winning yesterday's war.
By the metrics the board tracked, Intercom was succeeding. Revenue had roughly doubled to $200 million ARR over the prior two years under Karen Peacock.
The trajectory was the problem. Net new ARR had declined for five consecutive quarters and was approaching zero. Total revenue was still growing, but the growth engine had stalled out.
Six weeks after returning, he announced a second round of cuts that took another 13% of the company, with reductions hitting every department and geography. The cuts continued through attrition over the following year, and a parallel values rewrite forced out anyone who couldn't get behind the new direction. One Intercom executive later called it a "soft coup." Total turnover ran around 40%.
Then, one month after his return, ChatGPT launched. And McCabe saw the second, larger bet immediately.
From his own account: "When AI came. We started to build Fin, we realized that because the agent does the work that the humans do, the humans are not going to need the software they use to do the work. These agents were highly disruptive to our existing business. We saw it as a threat and an opportunity. We really didn't have any choice."
He pointed the entire product organization at the bet on top of the bet: that an AI agent, not software tools operated by human agents, would become the primary way Intercom's customers delivered customer service. The company committed nearly $100 million of its own cash to building Fin.
The execution was fast, though not as fast as Intercom's own marketing sometimes suggests. ChatGPT launched on November 30, 2022. Intercom's first features powered by GPT, built for human agents, shipped on January 31, 2023, roughly nine weeks later. Fin, the customer facing AI agent, entered public beta in March 2023 and launched to paying customers in June 2023, about seven months after ChatGPT.
Before Fin, Intercom's revenue was tied to seats. A seat was a customer service agent who logged in, triaged tickets, and responded to customers, and each one was a billable unit. When Fin began resolving conversations on its own (today it reaches a 51% resolution rate out of the box, climbs past 65% with tuning, and hits the 80s in the best deployments), the math changed.
If Fin resolves 60% of a customer's support volume, that customer needs 60% fewer human agents. Sixty percent fewer agents means 60% fewer seats. Sixty percent fewer seats, at the old pricing, means 60% less revenue.
McCabe's response was decisive. Rather than clinging to the seat model and bolting Fin on as a premium feature, he rebuilt the entire pricing architecture. Fin is priced at $0.99 per outcome, where an outcome is either a resolution, meaning the customer confirms Fin solved their issue or stops asking for help, or a procedure handoff, meaning Fin runs a configured workflow that ends in a resolution or an intentional handoff.
Customers still buy platform seats for their human agents on the Essential, Advanced, or Expert tiers, but the growth engine of the business now runs on consumption. The more work Fin does, the more the company earns. The seat revenue becomes a platform access fee, and the outcome revenue becomes the commercial center of gravity. Inside eighteen months, they had rebuilt their product, their culture, their headcount, their pricing model, and their commercial identity. They went from stalled growth to what McCabe described as faster growth than most public software companies, with over 300% growth on Fin specifically.
At Pioneer 2025, McCabe announced the next step. Fin would become not just a customer service agent but a full "Customer Agent" able to handle the entire customer lifecycle. In May 2026, the company renamed itself Fin, retiring the Intercom brand entirely. The product became the company.
Eighteen months later, HubSpot ran the same playbook from a very different starting position.
By the metrics that mattered to most boards, HubSpot in 2025 was a successful public company: about $3.1 billion in revenue, growing 19%, nearly 300,000 customers, and an inbound marketing method that had created its category. CEO Yamini Rangan and cofounder Dharmesh Shah had moved on AI earlier than most of their peers, pivoting the entire product roadmap in early 2023.
The trouble was that none of it was showing up in the stock. HubSpot shares fell by more than half over the year ending in early 2026, with a 24% drop in a single month at one point and a called-off Google acquisition adding to the volatility. The market was applying the same logic to HubSpot that it applied to every per seat SaaS company: if AI agents do the work, customers will need fewer seats, and HubSpot's entire revenue model is built on seats.
Rangan's response was iterative where McCabe's was violent. She moved HubSpot through staged repricing over two and a half years, culminating on April 2, 2026, when the company announced it was shifting its Breeze Customer Agent and Prospecting Agent to pricing based on outcomes, at $0.50 per resolved conversation and $1 per qualified lead. The transformation happened in three visible phases.
Phase 1: AI built in (2024). HubSpot launched Breeze in September 2024 as an AI layer across the platform: a conversational assistant, specialized agents for customer service and sales prospecting, and an intelligence engine. The decision to bundle it in rather than bolt it on was deliberate. Rangan's thesis, stated publicly across several earnings calls, was direct: "There should be one product, an AI-first product. We do not believe that there should be a separate hub or a separate add-on for AI." So Breeze was bundled straight into the Professional and Enterprise tiers, embedded into every hub. The AI was there, but the pricing model had not changed. Seats were still seats.
Phase 2: The credit system (2025). HubSpot introduced HubSpot Credits, a consumption currency that AI features drew from a shared pool. Professional plans included 5,000 credits per month, Enterprise got 10,000. Different AI actions cost different amounts: a Customer Agent conversation burned 100 credits, a prospecting research task burned 10. More credits could be bought at roughly $10 per 1,000. Credits are consumption pricing wearing a disguise. They break the link between revenue and seat count, which is necessary, but they are opaque. Customers cannot intuitively connect "100 credits" to business value the way they can connect "$0.50 per resolved conversation."
Phase 3: Pricing based on outcomes (April 2026). HubSpot made the leap. Starting April 14, 2026, the Breeze Customer Agent charges $0.50 per resolved conversation. The Prospecting Agent charges $1 per lead recommended for outreach. Not per interaction. Not per credit. Per result. You pay when the AI delivers a measurable business outcome. If it does not deliver, you do not pay.
This is the same destination Fin reached, arrived at by a different path. And the pricing tells an interesting story: HubSpot's $0.50 per resolution undercuts Fin's $0.99 by half. The price competition for outcomes delivered by AI has already begun. For every SaaS company still clinging to per seat pricing, this should be the alarm bell. The market is not waiting for you to decide. It is setting prices without you. The HubSpot path matters because it shows the transformation does not require McCabe's level of founder authority. It requires a willingness to iterate in public, to launch an imperfect model in credits and then improve it into outcomes, and to keep moving forward even when each stage draws criticism.
Two companies. Two hail marys. The same playbook. The same message to the rest of the industry: move now or get moved.
This chapter is for the companies that need to do what McCabe and HubSpot did.
Join companies like Zoom, DocuSign, and Twilio using our systematic pricing approach to increase revenue by 12-40% year-over-year.