The Marketing Millennials — Ep. 348

"Building an AI-First Marketing Mindset with Ryan Gavin, CMO at Slack"
Date: September 12, 2025
Host: Daniel Murray
Guest: Ryan Gavin (CMO, Slack)

Episode Overview

In this episode, Daniel Murray sits down with Ryan Gavin, CMO at Slack, to unpack how marketing leaders should be thinking about and integrating AI and agentic tools (like Agent Force) in their daily work. Ryan brings insights from his extensive experience at Microsoft, AWS, startups, and now, Slack within Salesforce. The episode is packed with actionable strategies, real-world stories, and perspective shifts for marketers ready to rethink their stack, workflows, and team structure in the age of AI.

Key Discussion Points and Insights

1. The Paradigm Shift: AI Will Transform Work

Timestamps: 01:34 – 04:04

• Change is accelerating:
  • Ryan emphasizes that the next five years will see more dramatic change than the last 25.
• Beginner’s mindset is a necessity:
  • Marketers must ask: “If I was starting this company today, in the age of AI, what would I do differently?”
  • The importance of reimagining ALL processes (email automation, funnel optimization, etc.) with an AI-first approach.
• AI adoption attitudes vary:
  • Some view AI as a superpower, while others feel secretive or anxious about leaning on these tools.
  • "Everyone approaches AI with a different mindset... but it is going to be pretty transformative. And there’s a lot of transformation that has to happen in marketing we’re barely scratching the surface of." — Ryan (03:29)

2. Productivity Myths & The Real Impact of AI

Timestamps: 04:04 – 05:45

• AI isn’t about laziness, it’s about leverage and output:
  • Increased tool adoption hasn’t shortened workdays but intensified expectations and output.
  • "It’s not like anyone’s going home earlier... The question is, you’re getting more done. Ideally, higher value work." — Ryan (05:09)

3. Practical AI at Slack: Onboarding, Knowledge, and Agents

Timestamps: 05:53 – 09:01

• Slack AI as onboarding superpower:
  • Ryan’s own onboarding experience: using Slack AI to unlock company memory, query strategic history, and get up to speed fast.
  • Real-world example: "I was rocking up to Slack AI like 10, 12 times a day just to get insights and understand… those are the types of things where it’s one example of AI able to tap into this long-term memory…" — Ryan (08:25)
• Agents as productivity multipliers:
  • AI 'agents' act like extra workforce, automating tasks previously requiring expensive human attention.

4. Agents vs. AI: Defining, Deploying, and Scaling

Timestamps: 09:01 – 13:05

• Clarifying terminology:
  • AI = capabilities (e.g., generate text or summarize).
  • Agentic AI/Agents = take action on your behalf with autonomy and specialization (“like teammates”).
• Case Study: Engineering Agent in Slack:
  • An agent runs in Slack, wires up knowledge, answers product queries for sales/customer success — handled 18,000+ questions at 76% accuracy.
  • Estimated time saved: over $20M in engineering hours.
  • "Not a lot of $20 million ideas just walking around... but these are the kinds of things we’re seeing." — Ryan (12:29)

5. AI-First Marketing Leadership: Where to Begin

Timestamps: 13:05 – 16:44

• Optimizing legacy activities:
  • Even "boring" marketing (like email) gets supercharged with AI for better send, copy, timing, and results.
• Agent-first customer journeys:
  • Predicting a future where prospects may never see your website — all buying happens via AI agents chatting with each other.
  • Marketers should consider: “What happens when discovery and decision-making happens entirely behind LLMs and agents?”

6. Rethinking SEO for the Agentic Era

Timestamps: 17:31 – 20:23

• Classic SEO still matters — but LLM optimization is the frontier:
  • Longer-form, well-indexed, comprehensive answers win.
  • "Anyone who says it’s not a black box is lying... but you can look at the patterns, and try to optimize for those patterns." — Ryan (18:23)
• Lead gen & qualification:
  • Marketers are seeing massive gains (e.g., 500% improvement at Box) by routing leads, running qualification, and even replacing portions of SDR work with agents.

7. First-Party Data as a Competitive Advantage

Timestamps: 21:13 – 24:25

• Activating organizational memory:
  • Connecting Slack conversations (context) with structured CRM data using features like Salesforce Channels.
  • AI agents prepare rich customer briefs — combining historical narratives and opportunity data with zero human time.
  • "Honestly — no offense to my sales teams — it’s way better than the briefing docs I used to get from sellers." — Ryan (23:28)

8. Data Hygiene in an AI/Agent World

Timestamps: 24:25 – 28:12

• Garbage in, garbage out still applies:
  • Instead of forcing compliance, focus on making it “brain-dead easy” for teams to log and share important info.
• Vision for seamless data updates:
  • Demo scenario: conversationally tell an agent what happened in a customer meeting, and the system updates CRM, schedules follow-ups, and notifies stakeholders automatically.

9. Reducing the Work of Work

Timestamps: 28:12 – 29:57

• AI can eliminate up to 40% of wasted time:
  • Manual entry, scheduling, data lookups, and context switching are all minimized — freeing teams for higher-value work.
  • "Slack serves as this command center... Then I bring agents to my people. I’m not asking people to go somewhere else." — Ryan (29:39)

10. Ryan’s Top 3 Slack Tips for Marketing Teams

Timestamps: 29:57 – 37:51

• Channels are the backbone:
  • Orient all initiatives, projects, and planning around channels. Use Canvas/docs for context.
• Embrace workflows:
  • Automate repetitive processes and use AI-generated workflows to further boost productivity.
• Leverage agents:
  • Use agents as virtual teammates — add them to channels, assign tasks, and let them execute.

"Every single employee gets a capability upgrade... every employee who wants to manage a team, turns out, you can — you’ve got a whole team of agents." — Ryan (37:22)

11. AI as an Employee Happiness Engine

Timestamps: 37:51 – 39:08

• AI and agents can handle least-liked tasks:
  • Documentation and process flows become almost automated — improving job satisfaction and engagement.

12. Human Judgment Still Matters

Timestamps: 39:08 – 44:31

• AI needs human orchestration and criticism:
  • Don’t “turn your brain off.” All work created by agents should be inspected, curated, and approved by a human.
• Team process:
  • Require marketing team to use LLMs for QA — have the AI grade or benchmark content before submitting for review.

"It's your job to question the work that's coming back to you. How are you going to judge the quality? ...We all become orchestrators of work." — Ryan (42:31)

13. The Marketing/Product Unity Principle

Timestamps: 45:11 – 48:45

• Hill to die on:
  • “Marketing is the product and the product is the marketing.” Deep integration is non-negotiable.
  • If marketing and product are siloed or disconnected, find another company.
• Interview audit tips:
  • Ask about recent in-product marketing tests.
  • Check if marketers are included in product reviews and engaged in experimentation culture.

Notable Quotes

• "Work's going to change more in the next five years than it has in the last 25."
  — Ryan Gavin (01:48)
• "It’s not like anyone’s going home any earlier ... Ideally, you’re doing higher value work."
  — Ryan Gavin (05:09)
• "AI agents become incredibly powerful... We saw it save over a quarter of a million engineering hours."
  — Ryan Gavin (12:29)
• "The traditional optimization of how do you get someone to your website… all that’s getting passed through. So what would your website look like if it was just agent-based?"
  — Ryan Gavin (15:13)
• "Every single employee gets to get this capability upgrade. And that’s the cool thing for marketers right now."
  — Ryan Gavin (37:22)
• "We absolutely cannot turn our brains off and just assume that what those digital teammates created is right."
  — Ryan Gavin (40:44)
• "Marketing is the product, and the product is the marketing."
  — Ryan Gavin (45:22)

Memorable Moments

• Ryan used Slack AI 10-12 times a day during onboarding to rapidly absorb company culture and knowledge. (08:07)
• The engineering agent at Salesforce/Slack saved $20M in engineering hours, answering 18,000+ questions in 6 months. (12:45)
• Ryan expects his team to use LLMs to grade/QA content before it comes to him for review. (43:09)
• If a company doesn’t have marketing integrated with product, "that would be the first thing I would be like, 'Hey, cool, that’s not the place for me.'" (45:36)

Important Timestamps & Segments

• [01:34] – Ryan on the pace of change & beginner's mindset
• [05:53] – Concrete examples of Slack AI in marketing onboarding
• [09:19] – Difference between AI and agents, with Salesforce examples
• [13:43] – AI for classic marketing activities & the agent-first web
• [17:51] – SEO in the age of LLMs, and the black-box challenge
• [22:10] – Using internal company memory and Slack for customer briefs
• [24:25] – Data quality & agent-powered automation
• [29:57] – Top 3 Slack structuring tips for marketing teams
• [39:08] – Human oversight in an AI-first workflow
• [45:11] – Marketing’s unity with product: Ryan’s hill to die on

Final Takeaways

• AI and agentic tools are not just about making marketing faster and leaner, but fundamentally reimagining the way work gets done.
• Marketers must lead with curiosity, be orchestrators (not just doers), and insist on tight integration with product.
• Organizations that help employees leverage company memory and agents will move faster and outstrip those who don’t.
• The brightest marketing talent will be those who continually reimagine their role and process for an AI/agent-first world.