- Product-Led SEO author Eli Schwartz has publicly argued that agencies telling clients to skip technical SEO in an AEO world are giving them bad advice.
- His argument is that crawlability, structured data and site architecture are still the foundations that allow AI systems to discover, understand and reuse content.
Agencies are selling a new AEO pitch: technical SEO matters less now that AI search is changing how people discover brands.
Eli Schwartz thinks that advice is wrong.
The growth advisor and Product-Led SEO author has stepped into a growing agency debate, saying the idea that technical SEO is no longer required for AEO is “completely wrong.”
His argument is simple: a site that AI crawlers cannot properly access, parse or understand can still lose visibility, no matter how strong the content looks to a human reader.
AEO may change the surface of search, but Schwartz’s point is that it does not remove the technical foundation underneath it.
Technical SEO is no longer required.
After spending the first few years of my consulting practice hearing from prospective clients that agencies told them they MUST begin an SEO engagement with an SEO audit, I am now hearing the exact opposite.
As a part of the whole SEO is…
— Eli Schwartz (@5le) June 19, 2026
On June 19, Schwartz took the argument public on X, pointing followers to a piece on his Substack, The Future of SEO & AEO, where he laid out the full reasoning.
The Agency Pitch That Started the Fight
What Schwartz is describing is a full reversal of a long-standing agency habit.
For years, SEO agencies often pushed technical audits as the mandatory first step of any engagement. Crawlability, site architecture, indexation, schema, JavaScript rendering and internal linking all became part of the standard starting point.
Now, according to Schwartz, some agencies are leaning on the rise of AEO to make the opposite case.
The pitch goes something like this: AI search has changed the game, so clients no longer need to worry as much about technical SEO. The old infrastructure work can be skipped. The focus should move to AI visibility, brand mentions and answer engine optimization instead.
Schwartz kept his response short: “This is completely wrong.”
The pushback from other practitioners came quickly. Local SEO specialist Harun Mia noted that technical SEO still matters as a foundation, because even strong content can struggle if the technical setup is weak. Developer and consultant Chaminda Delpagodage made a similar point in plainer terms: the cleaner a site’s plumbing is, the easier it is to cite.
What Schwartz Is Actually Arguing
The core of Schwartz’s Substack piece comes down to how AI systems discover and interpret content.
LLMs and AI-powered search systems do not magically understand a website because the writing is good. They still need to access pages, follow structure, identify entities, interpret relationships and extract useful information.
That is where technical SEO still matters.
Crawlability, structured data, internal linking and site architecture help search systems understand what exists on a site and how different pages connect. A site running on messy JavaScript, broken navigation or unclear architecture can still create problems for AI retrieval systems.
The implication is blunt: a page that cannot be crawled and parsed cleanly may as well be invisible to the systems that are supposed to reuse it.
That does not mean every site needs a giant technical audit before doing anything else. It does mean technical discipline is still part of the entry ticket.
Schwartz has made this point across multiple newsletter editions: AEO extends SEO, it does not replace it.
Where the Argument Gets Twisted
The confusion comes from treating AEO as either a total replacement for SEO or just SEO with a new name.
Both readings miss the point.
Schwartz has also argued that brand authority, off-site reputation and external mentions can shape LLM visibility in ways that traditional technical SEO cannot solve by itself. That part matters. AEO is not only about cleaning up crawl paths and adding schema.
But some agency pitches seem to jump from that point to the wrong conclusion.
If brand signals matter more, the argument goes, technical SEO matters less.
That is the leap Schwartz is rejecting.
Technical SEO is not the whole picture. It is the floor. What has changed is the structure being built on top of it.
Kristine, a technical SEO practitioner who responded in the thread, made that point clearly: technical SEO may be more important now because the LLM and agentic layers introduce new requirements on top of traditional search infrastructure.
That framing also connects to something Pedro Dias spotted inside Google’s own tooling. Google added an experimental “Agentic Browsing” metric inside PageSpeed Insights, with llms.txt listed as one of its signals.
As we covered earlier this year, Google’s messaging around llms.txt has become a live debate in the SEO community, especially now that the spec is appearing inside tools site owners actually use.
Why This Pitch Keeps Selling
There is a straightforward commercial reason the “no technical SEO needed” pitch works.
It takes a messy transition and makes it feel clean.
Clients hear a simple story: the old SEO playbook is dead, the new AEO playbook is here and the agency has the new thing they need to buy.
That kind of positioning works in a sales room. It is much easier to sell a clean break than a more complicated message about foundations, brand signals, content structure, crawling, citations and measurement all needing to work together.
The problem is that the new playbook still runs on much of the old infrastructure.
Hami Tahm, who responded in the thread, pointed to a different failure mode: brands spending months on audits while never asking whether AI systems can actually understand or trust their content.
That is a real problem.
But it argues for better technical work, not for removing technical work from the process entirely.
What Marketers Should Ask Agencies Now
For founders and marketing teams sitting across from AEO proposals, the sharper question is not whether an agency has an “AI visibility” product.
The sharper question is whether the agency can explain what happens when AI systems cannot properly access or understand the site.
Before accepting the idea that technical SEO is no longer required, marketers should ask a few basic questions:
- Can search engines and AI crawlers access the important pages?
- Can those pages be rendered and parsed cleanly?
- Is the internal linking structure clear enough to explain which pages matter?
- Does the site use structured data where it actually helps?
- Can the site clearly explain entities, services, locations and relationships?
- Are important pages hidden behind scripts, poor navigation or weak architecture?
If an agency cannot answer those questions clearly, its AEO pitch probably has a hole in the middle.
The broader takeaway is not that technical SEO should dominate every AI search strategy. It should not.
The takeaway is that AI search does not remove the need for clean technical foundations.
It makes weak foundations harder to hide.
