Episode Description

What happens when a top CRM vendor accidentally shifts from opt-in to opt-out on customer data training for AI? It triggers a "SASpocalypse" of user backlash. In this unscripted episode of CRM Playaz, hosts Brent Leary and Paul Greenberg dive deep into Hubspot's recent AI data policy controversy and its swift reversal. Joined later by executive advisor Esteban Kolsky, the Playaz introduce a brand-new segment—CRM Playaz vs. the LLMs—where they pit the consensus of major AI models against raw human experience to determine what's truly driving enterprise tech in 2026.

Show Notes

🎙️ The Hosts

• Brent Leary: Business analyst and technology speaker.
• Paul Greenberg: Author, strategy consultant, and the "Godfather of CRM."
• Special Guest: Esteban Kolsky, long-time enterprise tech analyst and executive advisor.

⏱️ Episode Timeline & Topics

00:00 – Warmups & Weather Banter

• Paul dials in from a highly humid environment, comparing his current weather to a late-summer night's talk show backdrop.
• Paul previews his upcoming trip to Newfoundland to celebrate his wife’s 80th birthday in her birthplace alongside 100 family members and childhood friends.
• The hosts trade stories about the high number of nonagenarians and octogenarians in their respective families.

04:30 – The Hubspot AI Data Policy U-Turn

• The Context: Brent shares a historic look at his relationship with Hubspot, displaying a vintage 2009 photo from the very first Inbound conference with co-founder Dharmesh Shah.
• The Controversy: The Playaz tackle a highly controversial policy change where Hubspot shifted customer data usage for AI model training to an opt-out mechanism rather than an explicit opt-in.
• The Reaction: Dissecting user outrage across platforms, Paul notes that for enterprise B2B software, customer data privacy is the holy grail. Charging high monthly user fees while quietly turning data over to enhance systemic models—where competitors might benefit—shatters foundational contract trust.
• The Reversal: Hubspot quickly listened to user backlash, apologized, and completely rolled back the policy change.

15:45 – Segment Debut: CRM Playaz vs. the LLMs

• Brent introduces a brand new experimental concept where they test a standard industry prompt across three major Large Language Models (ChatGPT, Claude, and Gemini) to see how their summaries match real-world reality.
• The Prompt: "What are the top three stories of the first half of 2026 when it comes to CRM and CX?"
• The LLM Outputs:
  • The models emphasize the "SASpocalypse" (Wall Street anxieties regarding seat-based software licensing versus AI agents), Salesforce's massive $3.6 billion acquisition of Finn, and the transition from systems of record to systems of autonomous action.
• The Human Reality Check: Esteban Kolsky breaks down why the LLMs are inherently echoing an unoriginal echo chamber. Because LLMs aggregate existing internet articles, they mistake vendor PR pushes and investor anxiety for what real executive buyers are actually doing.
• Esteban notes that at an executive board level, the frantic hype around agentic AI has cooled. Companies are no longer focused strictly on generative experiments; they are pivoting back to traditional foundational infrastructure, pricing realities, and finding sustainable, long-term multi-year growth.

💡 Key Takeaways From the Show

• Data Trust is Non-Negotiable: Consumer platforms like Meta can get away with broad opt-out data extraction because the services are free to end-users. Enterprise software platforms cannot bridge that same gap; software clients pay explicitly for secure siloed architecture.
• More Data Does Not Equal Better Models: As Esteban points out, cleaner, curated enterprise datasets produce useful strategic context, whereas simply harvesting mass quantities of unrefined customer inputs leads to chaotic, heavily commoditized outputs.

🔗 Connect with the Playaz

• Bookmark the show or suggest an AI prompt for the next round of Playaz vs. LLMs via our LinkedIn channels!