As a freelance SEO consultant who’s worked in-house, at an agency, and solo, I felt every bit of Patrick Stox’s career arc when we talked — he’s lived the same in-house/agency/freelance loop I have, and he’s just struck out on his own again. What I took away wasn’t a list of tactics; it was a reframe of what our job even is once AI systems start echoing what real people say about the businesses we represent.
“Go Fix the Product”: SEO as Part of GEO
Patrick’s most controversial take is that SEO is now part of GEO — and that the two aren’t the same thing. Influencing AI systems, he argues, is “more than SEO.” If an LLM or AI search engine is saying something wrong about your product, the highest-leverage move usually isn’t on-page work; it’s fixing the underlying product, because these systems echo what people genuinely complain about online.
He draws a clear line by company size: an SEO can still move the needle for a mom-and-pop local business, but a large brand can’t out-optimize public sentiment. “Big companies don’t have enough SEOs on their payroll to go change the opinion of the internet.” The work becomes business, branding, and product — not just keywords.
If you’ve got a problem with your product, the easiest way to get the LLM or the AI search system to say something different? Go fix that. Because if people are complaining on the internet, that’s what it’s gonna say. So it’s business, it’s branding, it’s so much more.
— Patrick Stox
Information Gain Is the New Frontier
With AI content generation now table stakes, Patrick says the frontier is making that content actually better. He’s candid that he’s rebuilding his own site with a roughly 160-page, AI-generated tech SEO hub — but it won’t stay that way. He’s adding summaries, checklists, a way for users to report inaccuracies, and plans to have friends mark up where he got things wrong.
The real unlock is information gain: sourcing your claims, and using AI to actually interview the experts you already have. Message five real experts inside your company a few times a week — over Slack, email, or even AI-driven calls — and capture their stories and experience instead of shipping unsourced generated text. “Anyone can do it, but it’s a matter of: can you do it well?”
Anyone can put out any number of pages. But if you’re just using it for content — just the content — you’re not providing resources, you’re not sourcing where you got something. So anyone can do it, but it’s a matter of: can you do it well?
— Patrick Stox
The Penguin Lesson AI Might Not Repeat
Patrick’s sharpest warning is about forgiveness. SEOs, he says, got lucky with Google: Penguin was painful for a few years, and then around 2016 Google effectively wiped the slate, forgiving bad links. The industry got a clean start that most never get.
AI search systems may not be so generous. Some of the tactics he sees companies scaling today look like obvious bad decisions — and the open question is how long these systems will “hold a grudge.” It might work now and hurt you for a long time later, so he urges caution about what you scale in pursuit of short-term wins.
SEOs got lucky with Google. We got Penguin, and for a few years that was really painful — and then they’re like, yeah, we forgive all your bad links. These AI search systems may make a different decision. How long will they hold a grudge against you?
— Patrick Stox
Can Indie SEO Tools Still Win?
Having helped build the product at Ahrefs for over six years, Patrick is realistic about competing with the two 800-pound gorillas. The giants are good for trust and reliability, but there’s still room for niche tools that focus, move faster, and do one thing better than the big platforms — he points to tools like Keyword Insights carving out their own space.
The constraints are data quality and upkeep. One person responsible for 10–15 tools won’t match whole product teams, and no indie can replicate Ahrefs-scale crawl infrastructure. His answer: go niche, lean on integrations and uploads, and accept that customers will keep their core tools (Ahrefs, SEMrush, Screaming Frog) while adopting yours for specific jobs.
Retrieved Isn’t Cited: Inside Google’s Machinery
Patrick’s depth on Google’s architecture is a highlight. He points to his public “How Search Works” deck (built while Ahrefs worked on Yep.com) and notes the data sources most people miss: it’s not just sitemaps and crawling, but RSS feeds, WebSub pings, GSC submissions, and the indexing API.
His most useful frame is retrieved vs. cited. In both Google and AI search, you might have 90 pages retrieved but only 13–30 cited. The low-hanging fruit of the AI era is being one of those retrieved-but-not-cited pages: you were relevant enough, but you didn’t say the right thing in the right way, so someone else got the citation.
What the Helpful Content Update Was Really Chasing
On the Helpful Content Update, Patrick’s read is that the extra “E” (experience) exists because too many SEOs were writing about things they don’t actually know. Polished writing only gets you so far; what the update rewards is hands-on, tested, original expertise — the obsessive testers like Project Farm, the pressure-washer flow-rate reviewer, or Cats.com sending food off for lab analysis.
He’s also clear that Google overcorrected. Some genuinely good sites got caught — he cites a knowledgeable running-shoe reviewer whose competitors were literally cutting shoes in half, and points to HouseFresh’s recovery. His hope is that Google reins it back, because even a site that simply aggregates information better is still a result he wants to see.
Connect with Patrick Stox
Patrick is consulting independently — reach him on LinkedIn or at patrickstox.com, and catch him at Tech SEO Connect in Durham this November.