What changes, and what stubbornly does not.

The two loudest positions are that SEO is dead and that nothing has changed. Both are wrong in ways that cost money: the first leads to abandoning things that still work, the second to missing a channel that is quietly growing.

What carries over unchanged

• Crawlability and rendering. If a system cannot access it, nothing else matters.
• Site speed and technical health.
• Genuinely useful content that answers a real question.
• Authority earned through citation and reference.
• Local fundamentals: accurate listings, real reviews, complete profiles.
• Structured data, which now does more work than it used to rather than less.

That is most of the discipline. Anyone telling you to start over is selling something.

Side by side

What genuinely differs

The unit changes from page to passage. You are optimizing sections that can be lifted out of context, not documents that get clicked.

Off-site weight rises. When a system synthesizes consensus, what others say about you matters more than what you say about yourself.

Measurement degrades. No rank, often no click, no referrer. You are reduced to sampling.

Being wrong becomes more expensive. Stale information gets restated confidently rather than merely sitting on a page nobody reads.

Where budget should actually move

Not away from SEO. Toward the parts of SEO that were always underfunded: entity consistency, structured data, third-party representation and measurement.

In most engagements the incremental cost of AI search work is smaller than the terminology implies, because the majority of it is work that should already have been happening. The genuinely new spend is monitoring, which is modest, and citation building, which is not.

How the platforms connect

It helps to know that the systems are not independent. Google's AI Overviews and AI Mode draw on Google's own index, and Google says there are no special requirements to appear in them beyond being eligible for Search. Copilot is grounded in Bing. ChatGPT search uses search partners and OpenAI's own crawler. Perplexity runs its own index and live retrieval.

So traditional SEO is the entry ticket for several of the biggest AI surfaces. What changes is what happens after a page is eligible: whether its content can be used in an answer, and whether the rest of the web agrees with what it says.

A budget example

For a local service business spending a fixed amount each month on SEO, a reasonable rebalancing looks like this: keep technical health, the Business Profile and the core service pages fully funded; reduce spending on generic informational blog posts; move that money into entity cleanup, review generation, a monthly AI monitoring routine, and outreach to the directories and publications assistants cite.

The total does not need to rise. What changes is the mix, away from volume and toward accuracy, corroboration and measurement.

Frequently asked questions

Is SEO dead because of AI search?

No. Traditional search still drives most discovery in most categories, and the fundamentals feed both systems. What has changed is the addition of a synthesis layer with its own requirements.

What is the biggest difference between SEO and AI search optimization?

The unit of optimization. SEO optimizes pages for position in a list; AI search optimizes passages for inclusion in a generated answer, which shifts weight onto extractability, corroboration and entity clarity.

Should I move budget from SEO to AEO?

Mostly you should move it within SEO, toward the parts that were always underfunded: entity consistency, structured data, third-party representation and measurement. The genuinely new spend is monitoring and citation building.

Does ranking on Google help with ChatGPT?

Indirectly. ChatGPT search uses its own crawler and search partners, so being indexed and well regarded across the web helps, but a Google ranking does not guarantee a ChatGPT citation.

Is AI search optimization more expensive than SEO?

Not necessarily. Most of the work is shared. The new costs are monitoring and outreach to non-link sources, often funded by reducing low-value content volume.