GEO is a reputation problem

by Miguel Lucas

Improving the image AI projects of you is not a matter for IT. Not keywords, not meta tags, not page load speed. In the era where machines have learned to speak human language, what they say about you depends, above all, on one thing: your reputation.

Language models do not form their opinion of a company by reading its corporate website. They form it the way a rigorous analyst would: by reading everything third parties have written about it. Press. Industry analysis. Reviews. Expert forums. A study by Omniscient Digital 1 of more than 23,000 AI-generated citations quantifies it: even when a user mentions a brand by name, only 23% of citations come from the corporate website. The remaining 77% come from editorial media, forums, review sites, directories, and content published by third parties, including competitors. Rewriting the corporate blog works on that 23%. Cultivating credibility with third parties works on the 77%.

The Princeton and IIT Delhi study that formalized the concept of GEO — Generative Engine Optimization, the discipline of optimizing brand presence in AI responses — 2 confirms this logic: the factors that most increase that visibility are not technical, they are editorial. Citing authoritative sources boosts visibility by up to 115% for lower-ranked sites 3. Including verifiable statistics raises it between 30% and 40% 2. Expert quotations, another 30–40% 2. None of this is code. All of this is reputation.

Here the irony appears. Communications and PR professionals have spent decades doing exactly what LLMs demand: building credibility through third parties, generating consistent narrative outside the perimeter controlled by the brand, positioning spokespeople in the sources that audiences — and now the algorithm — consider reliable. The vocabulary has changed: what was once earned media is now citation rate; what was once press coverage is now recommendation status. The mechanism is the same. The craft, fundamentally, is the same.

This is not just intuition. Gartner projects that by 2027 PR budgets will double as a direct consequence of this dynamic 4. Richard Edelman has dubbed it “The Golden Age of Earned” 5. And Ben Smith, co-founder of Semafor, puts it plainly: the best way to get your brand’s message into ChatGPT’s, Claude’s, or Gemini’s response is to talk to a journalist 6.

The most advanced information-processing technology we have ever built has decided that its favorite criterion for truth is the oldest one in existence: the opinion of a respected third party. Reputation professionals have spent years claiming a larger role in the conversation about digital. And it turns out that the LLM era is precisely the era in which their work is worth more. The irony is that many still don’t see them at the GEO table. And they are, precisely, the ones who should be chairing it.

Related theses

References

  1. Omniscient Digital — How LLMs Source Brand Information: An Analysis of 23,000+ AI Citations
  2. arXiv — GEO: Generative Engine Optimization
  3. Oomph, Inc — SEO vs. GEO Is a False Choice. Here's What Actually Matters.
  4. CMOtech Asia — Gartner: LLMs to reshape PR budgets & staff chatbots
  5. Edelman — The Golden Age of Earned
  6. Semafor — PR pros have discovered how to influence the chatbots: Talk to a journalist