Search Engine Land says GEO depends on reputation, not technical shortcuts

Why GEO is really a trust problem

The most useful thing in Search Engine Land’s latest GEO argument is how blunt it is: the winning question is not how to trick an AI system into noticing a page, but how to build a brand it can confidently describe. Gaetano DiNardi’s point is that generative engine optimization lives or dies on reputation, positioning, and third-party validation, not on isolated technical flourishes. If your brand is fuzzy in the market, AI search will reflect that fuzziness back at you.

That is a different way to think about visibility than old-school SEO, where a cleaner page, a better internal link path, or a more exact schema block could produce a visible lift. In this version of search, the machine is not just reading your site. It is checking whether your story makes sense everywhere else too, from expert mentions to category alignment to the language other sites use when they describe you.

The shortcuts everyone keeps selling

DiNardi’s critique lands because it targets the tactics that travel fastest on LinkedIn and X. AI info pages, markdown versions of content, automated llms.txt generation, FAQ stuffing, and superficial schema use all sound productive, but they are narrow fixes for a much broader problem. They can make a site look organized; they cannot make a brand look established, trusted, or well understood.

That is the real trap. A lot of GEO advice treats visibility like a formatting issue, as if the right file type or markup pattern will force recommendation behavior. Search Engine Land’s argument is that these tricks have marginal impact when the broader perception layer is weak. If a company is not already coherent in the market, the machine has little reason to treat its pages as the best answer.

What the reputation layer actually includes

The practical levers are less glamorous than a “GEO hack,” but they are the ones marketers can actually influence. Brand positioning, category definition, expert associations, reviews, third-party coverage, and consistent narrative framing all shape whether AI systems can confirm a company’s identity across the web.

That means the work sits at the intersection of PR, brand strategy, content, and search. If a company is described as one thing on its site, another way in analyst coverage, and a third way in customer reviews, the model sees noise. If the same category, value proposition, and proof points show up again and again in credible places, the model sees consensus.

Why AI search rewards consensus, not just pages

The technical backdrop matters here because it explains why reputation now affects discovery so directly. Google says AI Overviews and AI Mode may use a query fan-out approach, issuing multiple related searches across subtopics and data sources. Google also says AI Overviews use a customized Gemini model working alongside existing Search systems and the Knowledge Graph. That means the answer is being assembled from a wider evidence base than a single ranking page.

Google first brought AI Overviews to everyone in the U.S. and said people had already used them billions of times in Search Labs. More recently, Google has described AI Overviews as available in more than 120 countries and territories and in 11 languages. The footprint is already large enough that this is no longer an experimental layer sitting off to the side. It is part of the main path people take to information.

OpenAI has moved in the same direction with ChatGPT search, promising fast, timely answers with links to relevant web sources. Once search products start synthesizing answers and attaching citations, the brand story has to survive contact with multiple sources, not just your own site.

What the original GEO research got right

The idea of GEO did not come out of nowhere. The original paper from Princeton University, Georgia Tech, and the Allen Institute for AI was submitted in November 2023 and revised in June 2024. It framed Generative Engine Optimization as a black-box problem and reported that certain methods could improve visibility in generative engine responses by up to 40% in its tests.

That paper matters because it gave the field a vocabulary for thinking about generative visibility. But DiNardi’s article pushes the conversation forward by warning against reducing GEO to a checklist of technical tweaks. A 40% gain in a controlled test is not the same thing as durable brand preference in the wild. If the market does not know what you stand for, no amount of lightweight optimization will carry the story very far.

What marketers can control right now

The most useful takeaway is that GEO is not abstract. It is a reputation workflow with concrete inputs. You cannot force consensus, but you can make it easier for the web to agree on who you are and what bucket you belong in.

Focus on these practical signals:

• Tighten category language everywhere your brand appears, especially on your site, in bios, and in third-party profiles.

• Pursue expert associations that make your positioning legible, such as cited commentary, bylined articles, and respected mentions.

• Clean up review language and customer proof so the recurring themes match the story you want AI systems to learn.

• Keep product and company descriptions consistent across channels, because inconsistent phrasing creates weak signals.

• Use technical basics, but treat them as table stakes rather than the strategy itself.

That last point is important. Search Engine Land is not saying technical work is useless. It is saying technical work only gets you to the starting line. If the brand is not already credible, the page structure will not save it.

The new GEO playbook

The old temptation was to ask how to get a page crawled faster or interpreted more cleanly. The better question now is how to build a public identity that can be confirmed from the outside. AI search is built to synthesize, compare, and cite. That makes third-party validation, category coherence, and narrative consistency more valuable than clever shortcuts.

In that sense, GEO is becoming less like a search hack and more like a reputation system with ranking implications. Brands that already look coherent across the web will have an easier time being described favorably. Brands that rely on isolated technical tricks will keep wondering why the machine still does not trust them.