In this episode of the Unscripted SEO Podcast, we flipped the mic.
Jeremy Rivera ,the host you usually hear asking the questions , sat down with the NextBuild team as a guest. What came out of that conversation is one of the most practical breakdowns of how to actually integrate AI into an SEO and content workflow that we've featured on the show. If you haven't listened yet, we'd encourage you to do that first (link below). But if you want the highlights, here's what Jeremy covered.
Two Decades of SEO: The Long View on an Industry That Never Stands Still
Jeremy opened by laying out his background , 20 years across the full range of SEO practice: enterprise clients like HCA Healthcare (including an on-campus, security-supervised editing session), national restaurant chains, and local service businesses of every description. His point in sharing this wasn't to impress , it was to establish that he's seen the industry change enough times to recognize a real shift versus a false alarm.
His read on the current moment: the latest AI-driven transformation has had the biggest impact of any he's witnessed, and it's changed the nature of what counts as meaningful SEO work. The content volume era is effectively over as a competitive differentiator. Distribution and genuine human expertise are what matter now.
Commoditized Content and the Hive Mind Problem
One of the core arguments Jeremy made was about the 'tyranny of uniformity' in AI-generated content. Because the major LLMs are trained on largely overlapping datasets, asking them to produce content without additional context generates output that sounds similar regardless of brand, niche, or instruction. He watched this happen across client projects: different prompts, different clients, same voice.
He traced the roots of this back to content spinners in 2010 and early LLM-assisted content in 2015 , both tools that commoditized output before the current wave. Google's response to years of mass content production has been a de-emphasis on volume and a shift toward rewarding authentic, experience-based content. That's the EEAT framework in practice: experience, expertise, authority, and trust , qualities that a human conversation surface, and that a generic AI prompt cannot fake at scale.
The Content Pipeline Jeremy Built: Transcript In, Everywhere Out
The practical centerpiece of the episode was Jeremy's podcast-to-content workflow. Starting from a raw transcript, his process produces: a cleaned article for his agency site (SEO Arcade), a guest-facing recap, YouTube show notes, Castos distribution notes, and social media posts for LinkedIn, Facebook, Instagram, TikTok, Pinterest, Substack, and Medium , all from a single source conversation.
He's recently added Blotato, an MCP-connected social scheduling tool, so those posts go out automatically over a three-week window. Before that integration, he'd mention a new episode once and then move on to the next guest , a common pattern that leaves most of a podcast's distribution value on the table.
He also flagged a longer-term project: using Obsidian and a memory layer for Claude to enable cross-episode search across his 120+ interview archive. That's a corpus of roughly 3,000 words per conversation , an enormous amount of practitioner knowledge that currently sits 'locked in ice' as audio files.
Where AI Falls Short in SEO Analysis
Jeremy's answer to 'what can't AI do in SEO' was specific and useful: it can't hold meta-context across disparate data pools. The example he used was Google Search Console's sampled keyword data. For one of his pages, GSC only surfaces about 5% of actual keyword-level click data. A human SEO knows this and interprets the table accordingly. An LLM fed that table will draw the wrong conclusions , and do so confidently.
More broadly, 20 years of experience generates intuitions about data quality, tool limitations, and industry-specific bias that are genuinely difficult to encode into a prompt. That analyst instinct , knowing which numbers lie and which correlations are coincidental , is where the human layer remains irreplaceable.
The Distribution Shift and What It Means for Off-Site SEO
Jeremy referenced the emerging concept of distribution as the new content moat. Creating content at volume no longer guarantees that anyone sees it. The work now is in pushing outward: podcasts, social scheduling, PR, off-site brand mentions. But that external push has to have a center , a strong, well-structured home site that validates everything happening off-platform.
He also reframed link building in this context. The industry's fixation on high-DR backlinks caused practitioners to undervalue something that now matters more: getting the brand mentioned positively on 15 or 20 unique niche sites, with accurate descriptive language, in relevant context. That kind of multi-point brand presence is increasingly how LLM-based search systems build their understanding of what a company does and why it should be trusted.
On the "Zero Click" Question
Jeremy pushed back on 'the click is dead' framing. His reframe: it's not zero click, it's a second click in a second window. The user's journey is longer now ,they pass through AI-generated summaries before landing on a site , but the endpoint is the same. At some point, they still need to vet, verify, and purchase. That happens on a website.
The practical implication: make sure your site actually communicates what you do, to whom, why you're credible, and how to take the next step. Jeremy was pointed about how often this fails in practice , sites that don't list a phone number, don't state a clear service offering, and don't provide a specific value proposition. Those gaps were costly before AI search. They're disqualifying now.
Listen to the full episode at unscriptedseo.com. Subscribe wherever you get podcasts. Jeremy Rivera runs SEO Arcade (seoarcade.com) and the Unscripted SEO Podcast, and offers one-on-one consulting at jeremyriveraseo.com.