Lenny's PodcastEoghan McCabe: Why Intercom went wartime to bet $100M on AI
How a wartime culture rewrite drove 40% turnover and a soft coup; Fin, the AI agent, charges 99 cents per ticket and is racing past $100M ARR.
CHAPTERS
- 0:00 – 7:23
Intercom’s AI comeback: Fin’s growth and the urgency of disruption
Eoghan and Lenny open with the core claim: AI disruption is unavoidable, and Intercom’s bet on an AI agent (Fin) has reignited growth. They share headline outcomes—rapid ARR growth for Fin and strong competitive positioning—to frame the episode’s stakes.
- •AI disruption is framed as inevitable and aggressive for SaaS companies
- •Intercom shifts from late-stage SaaS to an AI-first, agent-led business
- •Fin is positioned as the future that will overtake the legacy product line
- •Early performance signals: rapid growth rates, large ARR trajectory, category leadership
- 7:23 – 10:06
What went wrong pre-pivot: plateauing growth, dilution, and a founder’s return
Eoghan describes Intercom’s pre-AI state: strategy drift, organizational bloat, and net new ARR sliding toward zero. He also shares personal context—illness, stepping down as CEO, and returning to stop the decline just before ChatGPT appeared.
- •Intercom’s strategy became unfocused (“all things for all people”)
- •Post-COVID “sugar rush” masked underlying problems
- •Net new ARR declined sequentially toward $0, nearing negative growth
- •Eoghan’s health issues and stepping away contributed to leadership transition
- •Returning as CEO was driven by pride and a desire to reverse the fade-out
- 10:06 – 12:34
The ‘all-in’ moment: GPT-3.5, an internal prototype, and ‘nothing to lose’ clarity
The turning point arrives with GPT-3.5: Intercom’s existing AI team quickly recognizes the step-change and builds an early Fin prototype within weeks. Eoghan emphasizes that bravery was aided by necessity—Intercom was already in a tough spot, making a bold pivot more viable.
- •Intercom already had rudimentary ML/bot efforts and an AI team in place
- •GPT-3.5 was immediately recognized internally as ‘different’ and disruptive
- •A working Fin beta emerged roughly six weeks after GPT-3.5’s launch
- •Intercom’s customer base and data scale created a unique advantage
- •The pivot felt urgent because the business was already under pressure
- 12:34 – 16:21
From ‘anti-bot’ to ‘pro-agent’: redefining ‘personal’ customer service
Lenny challenges Eoghan’s earlier reputation as anti-bot, given Intercom’s original mission to ‘make internet business personal.’ Eoghan reframes AI agents as more personal in practice—instant, consistent, expert help beats delayed human support with canned responses.
- •Intercom’s original mission conflicts on the surface with automation
- •Eoghan argues availability, expertise, and speed can be ‘more personal’
- •AI agents can deliver consistent, high-quality service 24/7
- •Humans remain valued, but AI is preferred for practical operational glue
- •Analogies include Waymo’s safety/reliability vs human variability
- 16:21 – 23:10
Repairing pricing trust: simplifying a meme-worthy model and refunding complexity
They revisit Intercom’s historically unpopular pricing, driven by an overly broad product strategy and value-capture across many metrics. Eoghan describes a deliberate simplification effort that reduced prices and effectively gave back large amounts of ARR to rebuild predictability and customer trust.
- •Pricing complexity stemmed from sprawling products and multiple value metrics
- •Unfocused strategy forced tiers, gates, and confusing charge units
- •A key leadership change: willingness to take short-term revenue pain
- •Intercom reduced prices and simplified terms, reportedly giving up major ARR
- •Better pricing improved retention and internal customer-obsession behavior
- 23:10 – 26:30
Outcome-based AI pricing: the 99-cent ‘ticket resolved’ bet (and early unit economics)
Eoghan explains Fin’s pricing philosophy: align revenue to outcomes, not costs, as a corrective to prior pricing scar tissue. Early on, costs exceeded price, but they set the price to what the market would value—and trusted costs would fall as the tech and stack improved.
- •Fin is priced per outcome: 99 cents per ticket resolved
- •Early economics were negative (cost > price) but expected to improve over time
- •Market research: many businesses spend $20–$30 per resolved ticket
- •Price selection balanced fairness, willingness-to-pay, and value signaling
- •Principle: pricing should follow customer value; cost is the vendor’s problem
- 26:30 – 28:32
Executing the transformation: cost cuts, lane selection, and a wartime posture
Eoghan details what ‘founder mode’ looked like: aggressive cost control, killing projects, and narrowing focus to customer service. He also describes making decisive calls without committee-driven processes, then allocating major capital to AI development.
- •Aggressively reduced burn and halted expensive, ‘old-world’ spending
- •Stopped or canceled multiple projects and initiatives to regain focus
- •Picked a strategic lane: customer service (especially as Zendesk weakened)
- •Shifted decision-making from consensus to decisive CEO calls
- •Reallocated capital heavily toward AI (near-$100M commitment mentioned)
- 28:32 – 31:19
Culture reset as strategy: rewriting values, performance scoring, and 40% turnover
A major part of the pivot was organizational: Eoghan rewrote company values to prioritize resilience, high standards, hard work, and shareholder value—then operationalized them with quarterly scoring tied to retention. This reshaped the company quickly but caused significant friction and turnover.
- •Values were rewritten explicitly to ‘cut out’ ineffective cultural patterns
- •Quarterly performance included both goal results and values-based behavior
- •A formulaic threshold triggered respectful exits to enforce standards
- •Turnover ultimately reached ~40% across a multi-year period
- •Employee sentiment reportedly swung to very high approval after the reset
- 31:19 – 39:52
Surviving internal backlash: the ‘soft coup’ and the case for strong hierarchy
Lenny presses on ‘soft coup’ references; Eoghan explains the resistance that arises when shifting from democratic/committee norms to top-down leadership. He argues great companies require clear hierarchy and CEO accountability, and that founder-led companies often outperform due to risk tolerance.
- •Major rule changes create friction—especially after ‘employee control’ eras
- •Eoghan argues CEOs must make unilateral, brave decisions with accountability
- •Founder-led companies are portrayed as structurally advantaged in disruption
- •Internal backlash included board-directed letters and organized resistance
- •The tradeoff: short-term pain to restore clarity, pace, and execution
- 39:52 – 45:12
Where agents go next: beyond CX to ‘agentic’ organizations and deflationary competition
The conversation broadens from Intercom to the future of business: customer experience is huge, but automation will spread to most repetitive operational work. Eoghan predicts flatter orgs blending humans and agents, with massive efficiency, deflationary pressure, and heightened competition.
- •CX is reframed as service, success, sales, and marketing—core business headcount
- •Repetitive operational work across functions will be automated (invoicing, HR flows)
- •Organizations will become human-agent hybrids with new oversight roles
- •Agents may exist at high levels (e.g., chief-of-staff-like coordination)
- •AI-driven efficiency will increase competition and lower prices for consumers
- 45:12 – 50:38
AI and jobs: replacing ‘demeaning’ work, reshaping sales, and preserving human trust
Eoghan takes a historical view: technology has always eliminated dangerous or repetitive jobs and generally improved quality of life, even if transitions are painful. He expects CX roles and many repetitive sales tasks (e.g., SDR work) to shrink, while human trust and connection remain valuable.
- •Job displacement is positioned as continuous with prior technological shifts
- •AI targets repetitive, mechanical digital work (e.g., macro-based support)
- •CX headcount likely declines; sales becomes more productive with fewer people
- •Human trust/connection remains central in complex selling and relationships
- •Acknowledges transition friction while arguing long-arc benefits to society
- 50:38 – 58:01
How to win in AI: ‘you don’t have a choice,’ talent density, and the youth-driven pace
Eoghan’s practical advice: embrace disruption directly, hire real AI talent, and empower younger builders who operate natively with AI tools. He warns that ‘sprinkling AI’ won’t work; competing with intense, AI-native startups may require radically higher speed and effort—or new leadership.
- •AI is framed as bigger than prior platform shifts (PC, internet, mobile)
- •Survival requires serious AI talent and leadership (e.g., Chief AI Officer role)
- •Younger teams use AI by default (vibe coding, AI-first workflows)
- •Competitive pace is extreme; partial measures and legacy culture won’t compete
- •Advice: if you won’t commit, hire someone who will and take a different role
- 58:01 – 1:04:35
Leadership and personal transformation: therapy, ego erosion, and staying human
Lenny notes Eoghan’s increased centeredness; Eoghan attributes it to years of startup adversity, long-term therapy/spiritual work, and the humbling experience of stepping away while sick and criticized. They discuss ‘ego death’ as more about smoothing edges than eliminating ego entirely.
- •Long startup tenure forces rapid personal growth through repeated adversity
- •Therapy is framed as performance-positive when done well (not ‘softening’ ambition)
- •Illness/time away and public criticism helped dismantle limiting ego identity
- •Ego doesn’t disappear; awareness reduces reactivity and defensiveness
- •Leadership benefit: clearer self-knowledge improves communication and authenticity
- 1:04:35 – 1:14:18
Why Intercom alumni become product leaders: autonomy, frameworks, and founder energy
Lenny shares research showing Intercom as a top producer of CPOs and founders. Eoghan credits a deeply product-centric culture, PM autonomy (mini-CEO ownership), and first-principles frameworks—plus historically hiring ‘founder types,’ even if they weren’t always ideal employees.
- •Founders drove product-first strategy; innovation/design were culturally central
- •PMs had substantial autonomy due to product sprawl and complexity
- •First-principles thinking and framework-building were taught and reinforced
- •Intercom popularized frameworks (e.g., RICE, Jobs-to-be-Done influence)
- •Hiring ‘founder types’ encouraged alumni to start companies and lead products
- 1:14:18 – 1:23:19
Lightning round and closing: recommendations, life motto, and a throwback to Quitter
They close with a lightning round covering books, movies, favorite products, and a ‘life is short’/memento mori philosophy. Eoghan also tells the story of selling his early Twitter app Quitter and how it taught him the visceral feel of strong product-market fit, then ends with a call to try Fin.
- •Reading/movie picks and reflections on craft/perfectionism (Fellow, Porsche 911)
- •Life motto centers on time passing quickly and living intentionally
- •Quitter origin story: built for himself, rapid adoption, sold later as a site asset
- •Call to action: try Fin at fin.ai and share with customer-ops leaders
- •Final wrap: Fin delivers value only when deployed to real customers (not ‘tire-kicking’)
