Every CMO planning next year’s budget is being asked some version of the same question: now that buyers ask AI assistants, should money move out of SEO and into generative engine optimization?
The framing is wrong, and the budget decisions that follow from it tend to be wrong too. GEO is not a replacement for SEO. It is a layer on top of it, and it should be funded that way.
GEO is a layer, not a replacement
CMOs should not move SEO budget into GEO wholesale. The foundations SEO pays for still feed AI answers, so cutting them to fund GEO removes the material GEO depends on.
The case for a straight swap usually rests on the idea that AI answers make rankings irrelevant. They change what a ranking is worth, certainly. But engines that search the web before answering still need pages they can crawl, understand and trust. The overlap and the differences are covered in AI visibility vs SEO. The budget implication is simple: protect the foundations, and fund the new work in addition.
Why do SEO foundations still matter for AI answers?
AI engines that retrieve from the web depend on the same things search engines do: pages that can be crawled, content that is clearly structured, and a site other sites refer to.
Crawlability
A page an engine cannot fetch cannot be cited. Technical SEO work on rendering, indexing and site structure pays off in both channels at once.
Clear, specific pages
Pricing, comparison and product pages that state facts plainly are easier to quote accurately than pages written around keywords. That is good SEO practice and good GEO practice.
Authority
The signals that make a page credible to a search engine overlap heavily with what makes it a plausible source for an answer. If your SEO program is weak, the first GEO dollar is often best spent fixing it. That is not a consolation prize; it is where the leverage is.
What does GEO add that SEO does not cover?
GEO adds four kinds of work most SEO programs do not fund: third-party source work, positioning clarity, accuracy correction and answer-level measurement.
Third-party source work
Answers lean on sources you do not own: review sites, directories, industry publications, community threads, analyst writeups. Getting represented well in those sources is closer to PR and partnerships than to on-page optimization, and SEO budgets rarely stretch to it.
Positioning clarity
An engine summarizing your category needs to know what you are, who you are for and how you differ. If your own site describes you three different ways, answers will pick one or skip you. Fixing that is product marketing work.
Accuracy correction
AI answers state things about your brand with confidence, and some of them will be wrong: old pricing, a missing feature, a competitor’s limitation attributed to you. Finding and correcting those claims at their source is ongoing work with no real equivalent in classic SEO.
Answer-level measurement
Rank tracking tells you where a page sits. It does not tell you whether an assistant named you, what it said, which sources it used or which competitors it named instead. Measuring that means tracking answers directly, per engine, on a fixed set of prompts.
How should you decide the split?
Decide it from evidence, in sequence: establish a baseline, fund measurement first, then move budget toward the engines and prompts where you are weakest. The right split is whatever the gaps justify, and it will differ by company.
- Start with a baseline. Find out how engines answer the questions your buyers ask today: which prompts name you, which name competitors instead, what sources get cited, and what gets said about you that is wrong. Without that, any split is a guess dressed up as a strategy.
- Fund measurement first. The first line of GEO budget should be the ability to see answers consistently over time. Measurement is cheap relative to the work it directs, and it is the only way to tell later whether anything else you funded made a difference.
- Move budget where engines split. The baseline will rarely show a uniform picture. You may be named reliably by one engine and absent from another, or strong on category prompts and invisible on comparisons. When engines disagree about you, the reasons are usually specific (a missing source, an outdated page, an unclear description) and fixable.
- Fund the teams that own the work. Third-party coverage belongs to PR and partnerships. Positioning belongs to product marketing. Accuracy fixes involve whoever owns the pages and listings in question. Route GEO budget entirely through search and the work that matters most goes unfunded, or gets done by the wrong people.
Then revisit the split as the baseline moves. When an engine that was ignoring you starts naming you reliably, the budget that fixed it can move to the next gap. A split that never changes is usually a sign nobody is reading the measurement it paid for.
Who should own the GEO budget?
GEO budget should be owned by the teams that do the work, not held entirely by the SEO team. The work that changes AI answers is spread across several functions, and the budget should follow it.
- SEO. Crawlability, site structure and the clarity of key pages. This is the foundation, and it usually stays in the existing SEO budget.
- PR and communications. Coverage in the publications and review sources engines cite, and getting inaccurate third-party descriptions corrected.
- Partnerships. Listings in marketplaces, directories and integration pages that describe what your product works with.
- Product marketing. One consistent description of what the product is, who it is for and how it differs, used everywhere engines might read it.
- Whoever owns measurement. The prompt set, the per-engine baseline and the reporting that tells every other team whether their work moved an answer.
A single owner can still coordinate the program. The point is that coordination and spending are different jobs, and routing every dollar through one team quietly decides which work gets done.
Some GEO spending is wasted
A GEO budget should not pay for rewriting the whole site for AI, for tools bought before a baseline exists, or for chasing individual answers day to day.
- Wholesale rewrites. Pages that are clear and specific already work. Rewrite the ones the baseline shows are failing, not everything.
- Tools before questions. Buying a platform before you know which prompts and engines matter produces dashboards nobody acts on.
- Daily chasing. Answers vary between runs. Budget that reacts to a single day’s result is budget spent on noise.
- Holding each channel to the other’s metrics. SEO has its own established measures and should keep them. Judge GEO spend on the answer layer: visibility, share of voice and accuracy on the prompts the work targeted, per engine, against the baseline.
What to do about it
In rough order of return on effort:
- Protect the SEO foundations: crawlability, clear product and pricing pages, authority.
- Fund a baseline of how each engine answers your buyers’ questions today.
- Fund ongoing answer-level measurement before any other GEO line.
- Fix the inaccurate claims the baseline finds, starting with pricing and capabilities.
- Give PR, partnerships and product marketing their own GEO budget for third-party sources and positioning.
- Move further budget toward the engines and prompts where the baseline shows you are weakest.
The AEO guide covers the answer-level work in more depth. The CMOs who get this right will not be the ones who moved the most budget. They will be the ones who can show, engine by engine, what the money changed.