Podcast Summary: Marketing Against The Grain

Episode: Why We’re Leaving ChatGPT for Google’s Gemini 3
Date: November 20, 2025
Hosts: Kipp Bodnar (HubSpot CMO) & Kieran Flanagan (HubSpot SVP of Marketing)

Overview

In this episode, Kipp and Kieran deliver an in-depth, excited analysis of Google’s Gemini 3, positioning it as the new standard for AI in marketing—surpassing even the latest from OpenAI’s ChatGPT (including GPT-5 and 5.1). They break down how Gemini 3’s advanced features—especially its video consumption, search transformation, and “vibe coding” capabilities—deliver practical, game-changing workflow improvements for marketers and entrepreneurs. Throughout, they share personal experiments, actionable strategies, and riff on how to monetize new AI-powered micro-apps.

Key Discussion Points & Insights

1. The Leap Ahead: Why Gemini 3 Outshines ChatGPT

• Scale and Model Issues with ChatGPT
  • ChatGPT 5 and 5.1 have routing issues: Users often don’t get the specific models they paid for, and overall, the experience feels less compelling than before. (01:11)
  • “It does not feel as compelling as Claude or Gemini’s latest model.” — Kieran, (02:19)
• Timing & Opportunity
  • ChatGPT’s recent limitations have opened a door for Gemini 3 to leap ahead, especially for marketers needing speed, intelligence, and multimedia capabilities.
  • “ChatGPT has kind of opened the door for other competition and Google’s walking right through it.” — Kipp, (02:28)

2. Multimodal AI Becomes Real: Video & Audio Understanding

• Gemini 3’s Multimodal Capabilities
  • Gemini 3 can analyze not just text, but also video and audio, delivering insights and actionable feedback.
  • Uses include: reviewing YouTube content, coaching podcast hosts, and curating video clips for content marketing.
  • “With advanced reasoning and the ability to analyze hour long videos, Gemini is perfect for students.” — Kipp quoting Gemini announcement, (14:41)
• Live Demo: Podcast Host Feedback
  • Kipp demonstrates uploading a podcast episode for critique, receiving detailed, actionable feedback on hosting technique and visual storytelling.
  • “All I did was that prompt and paste the link in and it gave us a detailed breakdown of our hosting.” — Kipp, (15:26)
  • Memorable Critique: “The issue: this episode was highly visual... you likely alienated your audio only listeners... The fix: narrate the action more explicitly.” — Feedback from Gemini, (16:44)

3. Gemini Transforms Google Search Into an App Platform

• Composability & Interactive Search

  • Google Search powered by Gemini 3 is evolving beyond blue links, delivering lightweight, interactive apps and custom UIs in response to queries (e.g., real-time data visualizations, explainer modules).
  • “It will actually return lightweight apps, not blue links.” — Kieran, (05:17)
  • Monetization Angle: These dynamic sidebars and apps can host targeted ads, offering new opportunities for marketers. (05:44)

• Victory Lap & Prediction Validation

  • The hosts reflect on their earlier predictions that search would become personalized, disposable software—a “microscope into the future of the Internet.” (06:32-07:10)

4. Vibe Coding: Productizing Your Prompts and Building Mini-Apps

• Hands-On Examples

  • Kieran shows how he used Gemini 3 to rapidly build:
    • A custom marketing strategy generator for HubSpot (analyzing competitor data, suggesting tactics, and prioritizing opportunities) — (07:48-10:27)
    • A customer intelligence tool to help startup founders identify and understand their ideal audience (extracting pains, buying signals, and active community locations from real web data) — (11:18-13:04)
  • Notable Productization Idea: “I’m going to create a Little app and charge $4.99 a report for a little test.” — Kieran, (10:34)

• Integration & Commercialization

  • Building, customizing, and selling AI-powered mini-apps is frictionless—except for payment integration, currently easier via Replit due to Google’s ongoing development.
  • “What really someone like me would like to do is just easily deploy that app and integrate payments... and just do it all from within Google.” — Kieran, (10:51)
  • Replit integration is a “feature flag” but present—promising future native monetization. (11:05)

5. Video Analysis & Social Content Automation

• Clip Extraction for Social

  • Gemini’s video analysis can select highly engaging, on-brand 30-second clips for channels like LinkedIn, complete with rationale and timestamps.
  • “It gives you the start and it gives you the end... That is actually really good. Holy smokes.” — Kieran, (19:14)

• Visual Frame Recognition

  • Capable of identifying the “most compelling frame” and explaining why, using visual theory and context.
  • Example: Analyzing an Ed Sheeran Heinz ketchup ad highlights color contrast and storytelling, not just the script.
  • “The hero reveal of the ketchup bottle... the bright saturated Heinz red cuts through the posh color palette... The cheap plastic industrial design of the bottle sits in direct opposition to the crystal chandeliers...” — Gemini via Kipp, (20:19-20:37)

• Generative Analysis for Ads

  • Gemini can generate complete ad breakdowns, identify hooks, psychological triggers, and even suggest net new concepts based on industry benchmarks.
  • “Imagine... find the best hundred ads of the year, find the best frame in each, and build a net new ad in VO3.” — Kipp, (21:13)

Notable Quotes & Moments

The Shift from ChatGPT

• “This is a step function. Better, I will say specifically from like a marketing and business growth perspective than the other models right now.” — Kipp, (21:49)

On Building Micro-apps for Marketers

• “I have so many $4.99 miniature apps I’m going to go and build this week.” — Kieran, (22:08)
• “It’s going to be Fiverr except a penny cheaper.” — Kieran riffs on 499-app pricing, (22:19)

Mic Drop on Video Analysis

• “It understands visuals to the level where it can pick out the frame.” — Kipp, (20:14)

Takeaway on Accessibility for Builders

• “It’s never been easier to build a company.” — Kipp, (12:41)

Key Timestamps (MM:SS)

• 01:11 — Challenges with ChatGPT 5/5.1 model routing
• 02:19 — Why Gemini & Claude now feel superior
• 03:30 – 07:38 — What composable AI in search really means
• 07:48 – 10:27 — Live demo: Gemini builds a marketing strategy app
• 11:18 – 13:04 — Customer intelligence with AI, from prompt to product
• 14:41 – 16:44 — Exclusive: Gemini’s video analysis and coaching
• 17:19 – 19:14 — Gemini curates and explains perfect social video clips
• 19:48 – 21:49 — Ad analysis, frame extraction, and future content workflows
• 22:08 – 22:19 — Kieran plans a “499 mini-apps” business using Gemini

Final Thoughts & Takeaways

• Gemini 3 sets a new bar for practical multimodal AI use in marketing, content creation, and small product generation.
• Marketers should prepare for a future where Google’s search is interactive and AI-assisted, not just information-based.
• The ease of building, customizing, and commercializing AI-driven micro-apps is about to create new SaaS models and revenue streams—especially for creators and niche experts.
• Stay tuned: Expect more deep dives as Kipp and Kieran continue to experiment with Gemini 3, driving new growth strategies from the front lines.

A must-listen for marketers looking to find the edge in the world of AI-driven content, apps, and strategy.