Jan 28, 2026 · 42m · saastr
Why Most SaaS Companies Will Fail at AI (And How to Avoid It) with Intercom's CPO
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this SaaStr Podcast keynote, Intercom Chief Product Officer Paul Adams outlines how legacy SaaS companies must dismantle traditional organizational silos, overhaul CRUD-based engineering paradigms, and replace seat-based monetization models with outcome-driven AI architectures to avoid obsolescence.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Jason holds 2.1% of the talking time here. How this is scored →
Jason as informed peer 0.0 Guest teaching 0.0 Guest disagreement 2.5 Jason pushing back 0.0
speaking balance: gold is Jason, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 11:27 The AI revolution cannot be negotiated
Adams delivers his most uncompromising stance, stating that AI transformation cannot be negotiated with reluctant staff and that legacy employees must be dismissed if they slow down execution.
Hardest push from Jason ▶ 1:41 Host event and discount promotionBecause this episode consists of a solo conference keynote with no conversational interviewer on stage, no host pushback occurs during the presentation.
Biggest teaching moment ▶ 15:17 SaaS seats make zero sense post-AIAdams dismantles conventional SaaS wisdom, explaining to the executive audience why traditional per-seat GUI pricing is obsolete compared to automated AI outcome pricing.
Jason holds their own ▶ 0:01 Host episode introductionThe host only participates during the opening podcast introduction and does not engage in dialogue or substantive technical exchanges during the recorded presentation.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Jason as informed peer | Guest teaching | Guest disagreement | Jason pushing back | Why |
|---|---|---|---|---|---|---|
| Intercom's Origin and Pivoting to AI | 0 | 0 | 2 | 0 | Paul Adams delivers a solo keynote presentation detailing Intercom's founding and its emergency pivot to bet the company on AI after ChatGPT launched. Because this is a monologue presentation without host interaction, host-side scores are zero. Adams adopts an assertive tone regarding the brutal nature of AI transformation. | |
| Instilling an AI-First Culture Without Sacred Cows | 0 | 0 | 4 | 0 | Adams forcefully argues that surviving AI requires non-negotiable culture change, noting that companies must be willing to part ways with a third of their staff if they resist fast-paced transformation. The host is not present on stage, keeping host scores at zero. | |
| Rebuilding Marketing and Deleting Seat-Based Models | 0 | 0 | 3 | 0 | Adams recounts blowing up Intercom's entire marketing organization upon taking it over and explains why traditional seat-based SaaS economics make no sense in an outcomes-driven AI world. Monologue format ensures zero host engagement scores. | |
| Defining Ambition and Deconstructing Workflows | 0 | 0 | 2 | 0 | Adams walks through the mathematical necessity of breaking down complex team workflows into discrete, highly reliable AI components to prevent compounding errors. Monologue presentation with zero host participation. | |
| Avoiding the Marketing Overhang and Fake Demos | 0 | 0 | 3 | 0 | Adams critiques the industry's widespread marketing overhang and fake demo culture, citing delayed enterprise launches like Apple Intelligence as cautionary tales. Monologue format with no host pushback. | |
| Software Development Shift: Empirical Evaluations and the Inverted Iceberg | 0 | 0 | 2 | 0 | Adams explains the inverted iceberg of AI development, detailing why Intercom abandoned its revered legacy design principles in favor of empirical evaluations and rigorous infrastructure testing. Solo keynote talk with zero host activity. | |
| Empowering Designers to Code and Doubling Engineering Velocity | 0 | 0 | 3 | 0 | Adams details how Intercom mandated that every product designer ship code directly to production and established non-negotiable targets to double engineering velocity. Keynote presentation with zero host scores. | |
| Navigating AI Go-to-Market and Complex Buyer Personas | 0 | 0 | 2 | 0 | Adams breaks down modern AI go-to-market dynamics, explaining that buying decisions have shifted from single functional managers to multi-stakeholder committees including C-suite leaders and AI technical evaluators. Monologue with zero host interaction. | |
| Operating Two Companies: Intercom vs. Fin | 0 | 0 | 1 | 0 | Adams discusses the operational and philosophical friction of running two distinct business models simultaneously: legacy Intercom seats versus Fin outcome-based consumption. Keynote monologue format. | |
| Critical Mistakes to Avoid in AI Transformation | 0 | 0 | 3 | 0 | Adams concludes by listing critical pitfalls SaaS companies face in AI transformation, emphasizing the danger of listening to legacy customers who resist change and refusing to make short-term revenue-harming decisions. Solo presentation with zero host scores. |
Statements from this episode (21)
Adams: Intercom experienced five quarters of declining revenue growth before ChatGPT
“We then had five quarters of declining revenue growth. So revenue is still growing, but the rate at which was growing was declining. And this was not a good period for the company. We had gone down the IPO journey.”
Adams: Intercom bet the company on AI two weeks after ChatGPT launched
“So we bet the entire company on AI. We bet the entire company. It was a risky thing to do. We ripped up our strategy. We ripped up our roadmap. We made these decisions in a very small number of days. Like you're talking once, two weeks after ChatGPT.”
Adams: Intercom's Fin resolves over 1M queries weekly across 6,000+ customers
“If you fast forward to today, Finn has over a million resolutions a week. So every single week Finn is resolving over one million customer problems. We have over 6000 Finn customers. And Finn's average resolution rate is over 65%. So the resolution rate over t…”
Adams: Pre-AI SaaS companies must overhaul every aspect of their business
“If you have a SaaS company and it predates AI and it's not an AI native company, in other words, you started the company designing an organization for the AI world, you have to change
Everything.
Your org, your product, your roadmap, how you build, the metrics…”
Adams: True AI companies must build model, AI, and app layers
“If you're a truly a deep AI company, of course you have an AI layer, but you also have an app layer and a model layer.”
Adams: Intercom's AI product Fin is close to $100M ARR
“You will grow to like a hundred million in AR or very quickly. Finn's on that path, very close to it.”
Adams: Companies must part ways with a third of staff for AI
“Like a third of your company, you need to part ways with them. They're not fit for this new world. They don't want to build the same way you want to build. They don't like fast, crazy fast pace. They don't like chaos.”
Adams: Employees needing constant convincing during AI pivots become liabilities
“Culture runs really deep. So you have to be wary of the people you need to keep convincing because they're going to be a liability.”
Adams: Traditional seat-based SaaS and CRUD GUIs make no sense post-AI
“Most SaaS companies sold seats. These are seats that humans sit in. Those humans use some kind of GUI, some CRUD type tasks to achieve some outcome. None of these things make sense anymore, right? None of these things make sense. There's no seats in a post AI …”
Adams: AI systems are easier to build when decomposing workflows discretely
“Getting AI and getting a rag system
to accurately do these things is easier when it's discrete. You can experiment much faster. You can solve bits of the problem independently and then kind of chain them together.”
Adams: Multi-step AI workflows suffer from compounding error rates
“All of these things compound. If you're really, really good at each of these steps, really good, like here we're really good, very reliable. You still have to multiply all these together. And by the time you're finished the entire workflow, especially the work…”
Adams: AI products require live production experimentation to verify they work
“When you build AI products, you really, really don't know if they're going to work. You need to have them in production, experimentation, all the time, trying to find edges, trying to find improvements in performance.”
Adams: In AI products, building UI is easy while infrastructure is hard
“AI products don't look like this at all. It's actually the inverse. Building the UI is the easy part. That's the small part. The hard part is building great AI infrastructure, building products that actually truly work at scale.”
Adams: Intercom abandoned its long-standing SaaS product principles for AI
“We had these principles at Intercom that evolved over years and they made us really good at building SaaS product... But, these principles were designed for the SaaS era. They were designed to build a SaaS product. We had to abandon all this shit.”
Adams: Intercom mandated all product designers ship code or leave the company
“We said every designer is now shipping code to production. It is now part of your job. If you don't like that, We had a great run, but we should part ways and you should go somewhere where designers don't have to ship code.”
Adams: Non-Technical Buyers Cannot Accurately Evaluate AI Products
“Customers don't know how to evaluate AI products. They don't know how to run good evaluations.”
Adams: AI Product Marketing Must Focus on Infrastructure, Not Features
“A lot of AI products do look the same. A lot of them are now AI agents. They all kind of look like some kind of chat interface. They all do different things, but they all kind of look the same. It's hard to tell them apart. Look at five competitors. They all k…”
Adams: AI Deals No Longer Have a Single Functional Buyer
“There isn't a single buyer anymore. So for example, with Finn, we have, yes, the customer service leader, and they're still heavily, you know, heavily influential. They have a lot of clout, but they don't make the decision. And the reason is because there's tw…”
Adams: Intercom's AI agent Fin is cannibalizing its legacy software usage
“Fin is totally different to Intercom. It's a completely different product. The two of them are like, they've almost nothing in common in some ways. It's a completely different product. I'm like, yes, Fin is, Fin usage is eating Intercom usage at times.”
Adams: AI Transformation Requires Making Painful, Revenue-Harming Decisions
“You won't make self-harming decisions to win. You'll protect revenue. You don't want to piss off the board. You know, you don't want to piss off your sales leader. Like there's loads of reasons and loads of places where you won't say, do you know what? We're g…”