Does GEO need its own budget, separate from SEO?

Crawlability, indexation, page speed, information architecture, internal linking and topical coverage all serve both disciplines identically. A page that cannot be crawled cannot be retrieved by an assistant either.

Content restructuring sits in an awkward middle. Writing an answer into the opening passage of a section is a genuine change in how you write, but it is done by the same people, on the same pages, in the same workflow. Splitting it into a separate budget line creates an argument about ownership without changing the work.

Google's AI surfaces make this concrete. AI Overviews are generated from pages already ranking, Gemini grounds in Google Search, and Copilot runs on the Bing index. Money moved out of conventional search to fund AI work weakens the foundation the AI surfaces are built on.

Both are new money rather than reallocated money. Both are also comparatively small next to a full SEO programme, which makes the conversation easier than people expect.

Measurement is the smaller of the two and should come first, because without a baseline you cannot tell which problem you have. Tooling for a single brand across the major assistants typically runs from a few hundred to a couple of thousand a month depending on prompt volume and how many competitors you track, plus the analyst time to actually read the answers.

The citation-sources programme is the larger line and scales with ambition rather than with company size. Review platform hygiene and profile work is a small fixed cost. A sustained community and publisher programme is a headcount conversation, because the work is slow, relationship-led and cannot be automated without becoming the kind of thing that gets you caught.

If your organic search performance is weak, GEO is premature. Fix the foundation first: it is cheaper, better understood, and it is the input to the largest AI surfaces.

If your category has very low assistant usage, which is still true in some industrial and public-sector-adjacent niches, measure once to confirm and then revisit in six months rather than funding a programme.

If you have never checked your crawler directives, do that before anything else. It costs an afternoon and it is occasionally the entire problem.