In the last year, almost every SEO agency has discovered "GEO". Most of them are doing what agencies always do when a new label appears: they have rebadged their existing service and put a new headline on the home page.
The capabilities they actually need to run GEO are mostly different from the ones they built for SEO.
Citation tracking is the first divide. Real GEO requires you to monitor, in real time, what AI engines say about your brand and your competitors across multiple surfaces. If your agency does not measure citation frequency or share of voice in AI answers, they are not doing GEO.
Briefing from citation gaps is the second. SEO briefs come from keyword volume. GEO briefs come from observed gaps — places where the engine names a rival or describes your category without anyone in it. If your agency builds content briefs from a keyword research tool, they are running the old playbook.
Off-site authority is the third. About 85 per cent of AI citations come from external domains. SEO agencies typically publish to your blog and call it a day. A serious GEO operator places content off-site, builds external authority and treats your owned blog as one channel among many.
Quality gating is the fourth. AI engines are increasingly hostile to thin, templated content. Real GEO programmes score every published piece — we hold a 100-point bar internally and refuse to publish below 80. Most SEO agencies still ship four to eight blog posts a month with no formal quality gate at all.
When you next talk to your agency, ask them four direct questions. How do you measure citations across AI engines? Where do your content briefs come from? What percentage of mentions you've earned in the last quarter come from external domains rather than your own site? What's your quality threshold and how is it enforced?
The agencies that answer those crisply are the ones building real GEO programmes. The ones who change the subject are the ones who rebadged the old work.