For two decades, brand visibility meant ranking. Customers and prospects typed a query, scanned a list of links and clicked on the first few. Entire marketing budgets were built around that behavior: keywords, backlinks, landing pages, paid search.
That behavior is eroding. A growing share of people now put their question to ChatGPT, Perplexity, Gemini, Copilot or Google’s AI Overviews, and receive a synthesized answer, often without clicking on anything. By 2027, we expect AI answer engines to be the main discovery channel for a large share of everyday decisions, whether people are choosing a car, a skincare brand, a bank, a software vendor or an employer. The consequences for communications budgets are more significant than most organizations have yet acknowledged.
From ranking to being cited
In a list of search results, being fifth still had value. In a generated answer, there is no page two. A brand is either named, described accurately and recommended, or it is absent. This applies to a consumer asking which running shoes suit a beginner as much as to a procurement team shortlisting suppliers. Generative Engine Optimization (GEO) is the discipline of becoming part of the answer.
Much of the early conversation about GEO has treated it as an extension of SEO: restructure your pages, add FAQs, adjust your schema markup. These steps are useful, but they address only the part of the problem a brand controls directly, and that part is the smaller one.
What the models trust
Large language models do not rank pages. They synthesize from sources they treat as authoritative. In practice, that means established news media, consumer and lifestyle titles, trade publications, expert commentary, reference sites, independent tests, product reviews and community forums. A brand’s own website is part of the mix, but it is weighed for what it is: one party’s claim about itself.
What makes a claim repeatable by an AI is third-party validation. When several credible, independent sources describe a brand in consistent terms (what it offers, what it is good at, who it is for, how it compares), the model has something solid to draw on. When those sources are missing, the model either ignores the brand or fills the gap with whatever it finds, including outdated or inaccurate material.
In other words, earned media is becoming the currency of visibility in AI answers. That is a fair description of what PR has always produced.
Why this is PR’s territory
The skills GEO requires are familiar to any communications professional: building a credible narrative, securing coverage in the outlets that matter to a given audience, positioning spokespeople as recognized experts and keeping messages consistent over time. What changes is the audience. Alongside journalists and readers, there is now a machine reader that aggregates everything published about a brand and draws its own conclusions.
SEO does not disappear. Technical hygiene, crawlability and structured content remain necessary conditions. But the scarce resource is shifting from keywords to credibility, and credibility cannot be engineered on a brand’s own domain.
What changes in practice
First, specialist media regain strategic value. For an AI engine answering a precise question, whether about electric motorcycles, family travel, skincare or industrial software, a respected specialist title, an independent review site or a regional outlet can carry more weight than a national headline. Audience size matters less than authority on the topic.
Second, consistency becomes measurable. Models aggregate across sources, markets and years. A brand that describes itself differently in each country, or whose positioning has shifted without the public record following, presents a blurred identity. Message discipline, long regarded as a soft virtue, now has direct consequences.
Third, substance is rewarded. Original data, clear expert positions and quotable, well-attributed statements are the material answer engines reuse. Announcements without information are not.
Fourth, measurement has to evolve. Coverage volume and share of voice remain relevant, but they need to be complemented by new indicators: how often a brand appears in AI answers to the questions its customers and prospects ask, which sources are cited, and whether the description is accurate. Regular audits of these answers should become as routine as media monitoring.
The international dimension
One point is often overlooked: AI answers vary by language and market. A query in French, German or Japanese draws heavily on sources in that language. A brand well represented in English-language answers can be invisible, or misdescribed, elsewhere.
For international organizations, this means GEO cannot be managed from a single market. It requires credible earned media in each country, built by people who know the local media landscape, its hierarchy and its expectations.
The budget conversation
Within the next budget cycle or two, most marketing and communications leaders, in consumer brands as much as in B2B companies, will be asked a simple question: which line funds our presence in AI answers? The honest response is that a large part of it sits in the PR line, and the organizations that recognize this early will have an advantage that is difficult to catch up on, because a credible public record takes time to build.
PR has long struggled to demonstrate its contribution to visibility in terms finance teams accept. GEO changes that. When discovery depends on what trusted third parties say about a brand, earned media is no longer a complement to the visibility strategy. It is the core of it.
You have a GEO and visibility project ? Contact us
