#24- AI Is Reshaping Agencies First: Building Products That Actually Work with Fergus Dyer-Smith

Join us for a thought-provoking conversation about AI transformation in the marketing and agency world with Fergus Dyer-Smith, Chief Product Officer at MSQ and CEO of M3 Labs. With a background rooted in product development, Fergus and his team build and deploy AI tools — including Assist, BrandCheck, Preflight, and Wave — for enterprise clients like WPP, Google, and Toyota. Fergus shares hard-won insights on what it actually takes to get AI working at scale inside large organisations.

In this episode, Fergus reveals why agencies were the first — not the last — sector to be disrupted by AI, and what that means for the future of the industry. He discusses why 90% of AI pilots fail and how companies can flip the script, the unique advantage mid-size agencies have over holding groups, and how human relationships remain the most defensible moat in an AI-driven world.

The Agency Paradox: Agencies are service-based businesses that many assumed AI would hit last — yet they are being disrupted first, creating both risk and enormous opportunity for those who move fast.

The 90% Pilot Failure Rate: Most pilots fail because companies ask "how do we get AI deployed?" instead of "what problems can AI solve?" The 10% that work are now driving budget increases across the board.

The Mid-Size Advantage: Large holding groups have firepower but lack agility. Mid-size agencies can move fast enough to deploy new tooling while still having the resources to compete.

The Human Moat: As AI output becomes commoditised, relationships, trust, and the ability to co-evolve with clients become the true differentiators — things AI cannot replace.

The Interface Question: From voice-to-text to agentic systems like OpenClaw, the very concept of how humans interact with technology is being redefined — and it could reshape entire business models.

Perfect for CMOs, CTOs, agency leaders, and business executives navigating AI transformation, particularly those managing large teams through cultural and technological change.

Fergus also shares his personal AI stack — including Claude, Cursor, Perplexity, Gemini, and Whisper Flow — and why staying on top of AI today means choosing a few trusted channels and committing to the pace.

"Humans are fundamentally lazy — they like to pass work off onto other people. But that's actually an opportunity: yes, we want to own our data and our rails, but can you come in and build it and manage it for us?" — Fergus Dyer-Smith