Everyone wants to talk about Generative Engine Optimization. We do too. But after running audit after audit, the honest pattern is this: most brands are not ready to be cited by AI because they were never set up to be found by conventional search either. Before GEO, the fundamentals are broken. Here is what we actually see across the sites we audit, in roughly the order it matters.
The page is not even indexed
The most common finding is the least glamorous one: pages that search engines and AI crawlers cannot see. A stray noindex left over from staging, a robots.txt disallow, a canonical tag pointing at the wrong URL, content that only renders after JavaScript runs. None of it shows up when a human loads the page, which is why it survives for months. But AI crawlers do not execute JavaScript, and a page that is not indexed cannot be retrieved, quoted, or recommended by anything. Indexing is not a detail you get to later. It is the gate everything else passes through, and it is the first thing we check.
There is nothing off-site to cite
The second pattern: the brand exists only on its own website. No directory listings, no marketplace presence, no earned mentions, few or no backlinks. This is a problem for conventional SEO and a bigger one for AI. Roughly 82% of AI citations come from earned media rather than a brand's own pages (Muck Rack), and across 1,885 pages studied by Ahrefs, brand mentions correlated about three times more strongly with AI visibility than backlinks did. If a company lives on one domain and nowhere else, the models have almost nothing to pull from when a buyer asks who to trust.
The content was published once and left to age
Freshness is underrated. We regularly find sites whose most recent update is two or three years old. Perplexity discounts pages older than about 90 days and cites recent content several times more often; AI crawlers also re-crawl on their own schedules, so a page that never changes slowly stops being fetched. Keeping content current is not busywork. It is what keeps a page in the set of things an engine is willing to quote this month.
Acquisition is measured through the wrong lens
This one is less technical and more expensive. A lot of brands have quietly decided that growth equals paid ads. Performance marketing gets glorified, budgets pour into it, and organic search plus the new layer of LLM search get underestimated, even though that is where a growing share of high-intent buyers now start. Underneath it, the cost of acquisition is often not clearly understood: the blended number hides how much a customer really costs once you strip out the traffic that would have converted anyway. Organic visibility and AI citation are compounding, owned channels. Paid is rented attention that stops the day you stop paying. Both have a place, but the ratio is usually wrong.
Chasing AI at the expense of the people
The last pattern is the one we are most careful about in our own work. It is possible to optimize a page so hard for machines that you make it worse for humans, walls of keyword-stuffed text, schema bolted onto a confusing layout, structure that reads well to a parser and badly to a person. That trade is a mistake. The same page has to convert the human who arrives from the answer. Optimizing for AEO and GEO should never compromise the UI and UX. When it does, you win the citation and lose the customer.
The order that works
None of this is exotic. It is the boring, foundational work that most sites skipped: make the page indexable, build a presence beyond your own domain, keep it fresh, measure acquisition honestly, and never break the experience for the human on the other end. Do that, and GEO stops being a separate project. It becomes the natural next layer on a site that was already built to be found. Skip it, and no amount of schema will get you into the answer.
If you want to know which of these apply to your site specifically, that is exactly what our free GEO audit checks, and the Approach page walks through how we sequence the fixes.