Intercom's AI pivot reveals brutal math: lower margins demand explosive growth

Mar 22, 2026 · Topline
🎧 PodShort 55 min squeezed to 2 Ai sprinklerAI Sales Tech
Episode artwork
AJ Bruno
CEO at QuotaPath
Assad Zaman
CEO at STA Sales Talent Agency
Sam Jacobs
CEO at Pavilion
Topline
55 min squeezed to 2
Full episode from Topline
Quotable Moments

You are trading a good business for a worse business. You have to grow at exponential rates as an agentic business because your gross margins are so much lower than SaaS.

Sam JacobsCEO at Pavilion
00:00

We have to burn the boats. The reason it's so hard is because they gave up $60 million of ARR. They had a multi-hundred-million-dollar business and this is why the innovator's dilemma exists.

Sam JacobsCEO at Pavilion
06:22

But I just think it's important because we see, you know, the famous VCs talking about massive growth rates and acceleration and it's good, but just understand 35% growth in a SaaS business is better than 35% growth on an agentic business because of gross margins. Just understand that.

Sam JacobsCEO at Pavilion
10:03
Key Insights
  • Trading a traditional SaaS business for an AI-native business requires exponential growth to compensate for significantly lower gross margins.
  • The venture capital community is largely ignoring the critical financial implications and higher cost structures of AI agent businesses compared to traditional SaaS.
  • Intercom executed a 'burn the boats' strategy by pivoting entirely to an AI-first approach with their agent 'Fin' after experiencing five consecutive quarters of zero net new ARR growth.
  • AI agents are fundamentally different from traditional SaaS, being compute-heavy and expensive to deliver, thus not achieving the same high gross margins.
  • Companies not willing to make decisive 'burn the boats' changes now will likely face an existential crisis by 2027.
  • Board support and a strong leadership vision are crucial for successfully navigating radical business transformations, especially when facing internal resistance or market skepticism.
  • The rise of AI will increase the demand for top-tier talent and computational resources, driving up both human and compute costs, contrary to some expectations of deflation.
  • In today's volatile market, growth is prioritized over margins because it enables access to capital, which is essential for continuous fighting, pivoting, and adapting.
Metrics Mentioned
  • Five straight quarters of zero net new ARR growth (Intercom experienced this before its AI pivot.)
  • $400 million in ARR (Intercom's current overall ARR.)
  • $100 million in ARR (The AI agent 'Fin' itself has achieved this ARR.)
  • $60 million of contracted seat-based ARR (Amount Intercom gave up by cutting legacy products during its pivot.)
  • $600 billion in market cap (Amount lost by SaaS companies in public markets over a six-month period.)
  • $300 billion in market cap (Portion of the $600 billion loss that occurred in just a few weeks.)
  • Single-digit growth (Intercom's growth before its AI pivot.)
  • Upwards of 35% growth (Intercom's current growth rate after the AI pivot.)
  • 90% gross margins (Typical for traditional SaaS businesses.)
  • $150/month for Claude Business vs. $200/month for Claude Personal (Illustrates pricing differences for AI services, with personal offering 4x usage.)
  • Low eight figures (Amount Intercom spent on the AI agent 'Fin' last year.)
  • $10,000 to $20,000 per month (Estimated inference costs per human for proactive AI agents.)
  • $60,000 annually (Estimated cost for an Intercom AI agent.)
  • 50% year-over-year growth (QuotaPath's current growth rate.)
  • 30% to 60% growth (Range for 'comfortable' growth rates for companies in today's market.)
  • 80% of Twitter's workforce cut (After Elon Musk acquired Twitter.)
  • 30 proper engineers (Number of engineers running Twitter after the cutbacks.)
  • 5% to 1% of people (The percentage of people who are 'the best' at any particular job.)
  • Hundreds of thousands of dollars of AI spend per month (Estimated AI spend for engineers.)
  • 10%, 20%, 30% inference costs (Inferred costs for a recruiter at STA in the next couple of years.)
  • 50% of Pavilion's customers (Are tech or tech-enabled, which is lower than the 80% three years ago.)

RevBots.ai View:

  • AI Sprinkler stage companies face a reckoning: bolted-on AI increases costs without transformation.
  • The ARM stage requires Intercom's 'burn the boats' mentality to rebuild GTM around AI economics.
  • Revenue leaders in SaaS Hoarder phase risk collapse if they don't address AI's margin compression.
  • Pavilion's tech customer shift (80% to 50%) signals broader market diversification beyond early AI adopters.
🎧Full Episode:Topline →