Embracing Marketing Mistakes

The episode contains useful tactical guidance on Copilot functionality (scheduled tasks, Co-work integration, Prompt Lab, token pricing at £85-145/month, marginal gains approach) and strategic points about GEO measurement and AI misinformation risks. However, these insights are scattered across meandering conversation with significant filler (opening jokes, side stories about the Forth Road Bridge, personal anecdotes), reducing overall density. The concrete tips on Copilot are valuable but standard best-practice territory rather than novel.

“the way to approach it is to do the list of things you want it to do, um, and then work out whether that's possible to get it to do it”

“it's 101 little things. It's not there's not kind of a big magic wand”

The core arguments are largely predictable: Copilot is underrated by critics, guardrails are strengths not weaknesses, AI isn't replacing SEO/PR but changing it, teams should move beyond individual tool use. The GEO discussion (who-what-why prompts, Google Analytics integration, monitoring) follows established frameworks. The misinformation stunt and negative SEO anecdotes are moderately interesting but not deeply counterintuitive. The episode reinforces existing orthodoxy rather than presenting fresh thinking.

“the guardrails are there is if you are working for a pharma company or government or financial services, you need them”

“the real power is when you use it as a team”

Stuart Bruce is a legitimate practitioner with genuine in-house AI implementation experience (Copilot, Co-work, Frontier Program), client work, and teaching credentials (Leeds PR degree). Darrell Grimes runs Hard Numbers and Prohibition, has hands-on GEO research and execution experience, and can cite real client outcomes. Both have credible operational depth, though neither is a C-suite exec or hyper-visible figure. They are practitioners-at-scale rather than pure thought leaders, which is appropriate for the show's premise.

“we organized an invite-only breakfast recently with Simon Boor”

“we've done a uh research report, it's out this week, hardnumbers.co.uk forward slash gullible”

The episode includes specific numbers (64% of CEOs, 13% using AI, 70% of junior team with no oversight; Copilot pricing £85-145/month; 85-90% accuracy on email sorting; 1.3 million injected pages; average cited source age 19 days; 60-90% traffic from ChatGPT in GEO analysis). However, these are often stated without supporting sources or methodology. The Copilot examples (breakfast scheduling, HubSpot integration) are concrete. The GEO discussion references 'research reports' without full detail. Compared to data-dense B2B content, this has moderate specificity but lacks depth on causation and measurement frameworks.

“Copilot is on tokens, and it's predicted tokens. The pricing model is on tokens. There's a pay as you go model. Yeah. So, and you can set caps by user, by by group, or by tenant”

“the average age of a cited source is something like 19 days, right”

The host (Chris Norton) asks reasonable follow-up questions and gently probes (e.g., 'Have you made any mistakes?', drilling into GEO measurement). However, the conversation often meanders into tangents (Forth Road Bridge reference, personal anecdotes, multiple false starts on naming clients) and lacks sharp pushback or disagreement. The guests are generally unchallenged; there are few moments of productive tension or clarifying interrogation. The dialogue feels collaborative and friendly but misses opportunities to stress-test claims or expose weak spots in reasoning.

“I'm intrigued because I've got a view on it. So give me your rant”

“so what are your hot tips on co-pilot then to how to get the most out of co-pilot”