<p>One of the things I keep coming back to in my consulting work is how the conversation with clients has fundamentally shifted. It used to be about rankings and traffic. Now it’s about survival strategies — how do you stay visible when the rules keep changing? I had one of the most grounded conversations I’ve had on this topic with <a href=”https://www.linkedin.com/in/joelelliottmiller”>Joel Miller</a>, who runs <a href=”https://www.theskyfloor.com/”>The SkyFloor</a> alongside his identical twin brother. They’ve been in the digital space since 2008, and what struck me most wasn’t their technical knowledge — it was how clearly they’ve thought through what <em>actually</em> creates durable client value in a market that keeps shifting beneath our feet.</p>
<p>🎙️ <strong><a href=”https://the-unscripted-seo-interview-podcast.castos.com/episodes/why-human-judgment-still-scales-joel-miller-on-ai-seo-and-owned-audiences”>Listen to the full conversation</a></strong></p>
<p>▶️ <strong><a href=”https://youtu.be/CQHM2Ml6QSc”>Watch on YouTube</a></strong></p>
<p>
</p>
<h2>Redefining What an Agency Actually Sells</h2>
<p>Joel doesn’t lead with services. He leads with outcomes. The SkyFloor calls itself a team of problem solvers — their job is to understand what a client is trying to become, then work backward to figure out what tools get them there. That’s an important distinction from selling deliverables.</p>
<p>He mapped it onto a pyramid I’ve been using in my own client conversations: the lowest tier is executing a specific task, the middle tier is saving the client time, and the highest tier is enabling them to become who or what they want to be. The higher up the pyramid your service sits, the harder it is to replace — with a cheaper vendor or with AI. The human judgment required to genuinely understand a client’s goals and translate them into strategy is still, as Joel put it, a distinctly human task.</p>
<h2>A Practical Example: What AI Can and Can’t Do in Sales Analysis</h2>
<p>Joel shared a case study I found genuinely useful. For a client with sales process problems, he listened to 30–40 hours of incoming lead calls himself — fully, by hand, taking notes. The AI could have analyzed the transcripts. But it would have missed the nuances that Joel was actually listening for: tone shifts, the weight of silence after a pricing objection, the difference between enthusiasm and polite deflection.</p>
<p>Once he’d mapped those patterns through human listening, he built a custom GPT using that framework — and then connected it to CallRail for automatic call grading. The model works because the human did the difficult perceptual work first. AI scaled the scoring. Human defined what to score.</p>
<p>That’s the practical lesson I take from it: AI is best deployed <strong>after</strong> a human has defined the problem clearly. It is not a substitute for the judgment needed to define the problem in the first place.</p>
<h2>The Click Collapse Is Real — and Vertical-Dependent</h2>
<p>Joel confirmed something I’ve been seeing across my own client work: AI Overviews are driving clicks down 40% or more for some content categories, even as impressions hold or grow. The SERP composition has changed — for many informational queries, organic results don’t appear above the fold at all. It’s AI Overview, then local pack, then ads, and organic shows up at the bottom if you’re lucky.</p>
<p>His practical response on the paid side: shift budget toward high-intent near-me combinations that still reliably trigger ad placements. On the organic side: stop relying on search volume data alone and start understanding the full SERP composition for each target query. You have to know what you’re actually competing against before you decide whether it’s worth trying.</p>
<h2>Owned Audience as the Durable Asset</h2>
<p>This is the part of the conversation I keep returning to. Joel is actively steering clients away from traffic-as-output toward <strong>audience-as-asset</strong>. The logic is simple: traffic derived from Google’s algorithm is rented. An email list, a subscriber base, a community you’ve built — that travels with you regardless of what Google does next.</p>
<p>The hard part isn’t the strategic argument. Clients generally agree with the logic. The hard part is the personality requirement. Building an owned audience — especially on LinkedIn — requires someone at the organization to be genuinely, consistently out front. Personal pages outperform business pages by roughly 8x. That means coaching clients into a level of personal visibility that many are uncomfortable with.</p>
<blockquote><p>”The human element is going to be the most important thing. It always was, but it gets highlighted when all these other barriers disappear.” — Joel Miller, The SkyFloor</p></blockquote>
<h2>The AI Slop Problem and What It Means for Authentic Brands</h2>
<p>Joel and I both landed on the same observation: the brands and creators who leaned hardest into AI-generated content at scale have unintentionally created an opportunity for the ones who didn’t. Gen Alpha is already using “that’s so AI” as an insult. Authenticity is the scarcest commodity in the current attention economy — not production quality, not volume, not speed.</p>
<p>Even as AI tools continue to improve — and they will — the question that marketing has always had to answer remains: why should anyone care, and how do they know the person behind this is worth listening to? AI can produce content. It cannot produce the credibility and relationship that makes content meaningful to an audience.</p>
<hr />
<p>🎙️ <strong>Full episode:</strong> <a href=”https://the-unscripted-seo-interview-podcast.castos.com/episodes/why-human-judgment-still-scales-joel-miller-on-ai-seo-and-owned-audiences”>Unscripted SEO Podcast — Joel Miller</a></p>
<p>▶️ <strong>Watch:</strong> <a href=”https://youtu.be/CQHM2Ml6QSc”>youtu.be/CQHM2Ml6QSc</a></p>
<p>Find Joel at <a href=”https://www.theskyfloor.com/”>theskyfloor.com</a> and on <a href=”https://www.linkedin.com/in/joelelliottmiller”>LinkedIn</a>.</p>