How AI is rewriting startup growth playbooks
- B2B
- AI, Marketing Operations
- Marketing Consultant
- Artificial Intelligence, Sales Enablement, Growth Marketing
Most AI pilots fail because founders layer features onto old playbooks instead of rebuilding revenue architecture. Mark Roberge, co-founder of Stage 2 Capital and former founding CRO at HubSpot who scaled from $0 to $100M revenue, explains how AI compresses traditional sales cycles and enables systematic growth. He breaks down the four-phase abstraction from AI-supported reps to full revenue automation, details how selling time can jump from 25% to 75% of a rep's week, and reveals why sustainable moats now depend on owning the point of work rather than systems of record.
- Part 1 How AI is rewriting startup growth playbooks
Episode Chapters
-
01:47: Most AI pilots fail
95% of Enterprise AI pilot programs fail to deliver meaningful revenue growth because founders are just paving cow paths instead of rebuilding revenue architecture around AI-first models.
-
08:53: When to scale calculations
The cause of u ecessary startup failure is usually the incorrect decision on when to scale revenue and how fast, which can now be calculated using your own data through the Science of Scaling methodology.
-
10:45: Selling time revolution coming
Average selling time today is 25% of a salesperson's week, but AI can easily triple productivity by pushing that to 75% within two years by eliminating administrative overhead.
-
11:33: ICP redefinition gets daily
Define your ideal customer profile based on where you create unique customer value, not lowest CAC, and AI enables this computationally intense exercise to happen daily instead of quarterly.
-
12:38: 22-year-old SDRs replaced
AI can analyze 57,000 companies in seconds to identify the right 120 accounts to target, eliminating the current reality where account selection decisions are made by inexperienced SDRs.
-
17:38: Temporary moats get crushed
If competitors can replicate your feature advantage in three months, it's a temporary moat that will get you killed when incumbents put 50 engineers in a room to reverse engineer your product.
-
19:42: Vertical AI wins compliance
Vertical AI companies have the strongest moats because CTOs at hospitals, banks, and law firms care more about HIPAA and SOX compliance than building on raw LLMs.
-
24:16: Own the point of work
The sustainable moat shifts from owning the system of record to owning where raw, unstructured data collection occurs - like conversation transcripts between buyers and sellers.
-
25:04: Fast followers beat first movers
Rigorous research shows first movers win only 12% of the time while fast followers win 88%, especially brutal in AI where first movers raised massive valuations and can't drop ACVs.
-
34:47: Scale on customer success
Scale when you have consistent, predictable customer success creation, not predictable revenue creation - that's where most founders mess up the timing.
-
Episode Summary
-
Why 95% of AI Sales Tools Are Building on Quicksand
The Brutal Reality
Tom Chavez opens with a gut punch: 95% of enterprise AI pilot programs are failing to deliver meaningful revenue growth. The culprit? Founders are "paving the cow path" - sprinkling AI on top of broken playbooks instead of rebuilding their revenue architecture from scratch. Mark Roberge, who scaled HubSpot from $0 to $100M as founding CRO, doesn't mince words about the current state of AI sales tools: "Everyone's building features looking for problems while the incumbents are about to eat their lunch." -
The conversation quickly exposes a harsh truth most VCs won't tell you. While founders chase the latest AI wrapper plays, they're missing the fundamental shift happening in go-to-market strategy. "The average salesperson spends only 25% of their time actually selling," Roberge reveals. "AI can flip that to 75% in the next two years - that's a 3x productivity gain just from eliminating busywork." But here's the kicker: most AI startups are building point solutions that nibble at the edges instead of attacking the core problem.
-
The Post-Check Playbook
Abstraction Layers: Your Roadmap to Relevance
Roberge lays out four phases of AI abstraction in sales that separate the wi ers from the walking dead. Phase one - where we are now - focuses on augmenting reps to maximize customer-facing time. Phase two replaces the seller entirely. Phase three eliminates the buyer's journey. Phase four? "Functional utopia where product, marketing, and sales blur into one continuous experience." -
The immediate opportunity lies in that first abstraction layer. Roberge walks through specific examples: AI that analyzes 57,000 companies in seconds to identify your next 120 targets, crafts personalized outreach for entire buying committees, and provides real-time coaching during sales calls. "A 22-year-old SDR making target account decisions is insane when AI can do it better in milliseconds," he states bluntly.
-
The Moat Reality Check
"What's your barrier to entry?" Roberge asks every founder. When they cite a feature, his follow-up is brutal: "Cool, so when Competitor A reverse-engineers it in three months, then what?" The distinction between temporary and sustainable moats has never been more critical in the AI era. -
Roberge identifies the real sustainable moats emerging in AI: owning the point of work instead of the system of record, vertical AI solutions with deep compliance requirements, and the fast-follower advantage. "I have more conviction in vertical AI plays right now," he admits. "The CTO of a hospital doesn't care about your LLM - they care about HIPAA compliance and proven workflows."
-
The Fast-Follower Advantage
Here's the contrarian take that makes VCs squirm: "Studies show first movers win only 12% of the time. The fast follower wins 88%." Roberge dismantles the blitzscale mythology with surgical precision. Those AI startups that raised at billion-dollar valuations on $10M revenue? They're trapped. They can't drop prices, can't pivot strategy, and have to grow into impossible expectations while fast followers build better products at 5% of the cost. -
"Google wasn't the first search engine. Salesforce wasn't the first CRM," Roberge reminds us. The fast follower advantages in AI are even more pronounced: the technology stabilizes, the category is already evangelized, and you can literally position as "Acme Company at 90% off."
-
The Unfair Advantage
The conversation crystallizes around three frameworks sophisticated founders can implement immediately. First, stop building features and start owning the raw data collection points - transcripts matter more than CRM entries. Second, focus on customer success creation, not revenue creation. "That's where people mess up," Roberge emphasizes. Third, embrace the fast-follower position if you're not already locked into an overvalued cap table. -
Roberge's parting wisdom cuts through the AI hype with characteristic clarity: "Creativity is the one skill AI won't replace. Everything else is just knowing what already exists and spitting out answers." For founders building in the age of AI, the message is clear: stop adding AI features to broken processes. Rebuild your entire go-to-market architecture around AI-first principles, or watch fast followers do it at a fraction of your burn rate.
-
Up Next:
-
Part 1How AI is rewriting startup growth playbooks
Most AI pilots fail because founders layer features onto old playbooks instead of rebuilding revenue architecture. Mark Roberge, co-founder of Stage 2 Capital and former founding CRO at HubSpot who scaled from $0 to $100M revenue, explains how AI compresses traditional sales cycles and enables systematic growth. He breaks down the four-phase abstraction from AI-supported reps to full revenue automation, details how selling time can jump from 25% to 75% of a rep's week, and reveals why sustainable moats now depend on owning the point of work rather than systems of record.