There are two loud camps right now. One says SEO is dead and AEO is a whole new discipline. The other says it's the same SEO with a new acronym slapped on to sell consulting.
Both are wrong, in the way that confident takes usually are. Here's the honest version, and what to actually do about it.
SEO and AEO are two stages of one game
When you ask ChatGPT or Perplexity a question, two things happen in order.
First, the engine retrieves a set of sources to answer from. To be in that set, you still need the classic search fundamentals: crawlable pages, clear topical coverage, useful answers, and authority. Google's AI features guidance does not say to throw SEO away. It says AI experiences still depend on eligible, understandable web content.
Second, out of everything it retrieved, the engine decides which sources to name. That is where citation signals matter: a clear entity, trusted third-party mentions, consistent language around what you do, and content that gives the model something specific to use. A recent CXL citation analysis is a good reminder that answer engines do not reward vague filler just because it is long.
Rankings get you retrieved. Semantic signals get you cited. That's the whole relationship between SEO and AEO in one line. You need both layers, and most brands are only running one.
What still matters (don't throw out the SEO playbook)
The "SEO is dead" crowd is about to make an expensive mistake. The things that get you retrieved are the same things SEO always rewarded:
- Ranking for the searches buyers actually make.
- A fast, crawlable, well-structured site.
- Real authority and credible mentions.
- Content that genuinely answers the question.
If you stop doing this, you fall out of the retrieval set entirely, and no amount of "AEO" saves a brand the model never pulls in the first place.
What is genuinely new
Here's where the AEO crowd has a point. A few things really did change.
- The old playbook → The new playbook
- Rank for keywords → Get cited in the AI's answer
- Optimize the headline → Optimize the answer inside the content
- Track CTR and position → Track citation share and zero-click visibility
- Earn backlinks → Earn brand mentions in trusted context
- Win the click → Win the recommendation, click or not
The deepest shift: you're no longer only optimizing a page. You're shaping what the model believes about your brand. AI doesn't rank you, it remembers you, and that memory is built from what the trusted sources across the internet say about you, not from your homepage. The work moves from rank tracking toward something closer to memory engineering.
"The 'is SEO dead' debate is a distraction. Retrieval is still SEO. Citation is the new part. You don't get to skip either one, and the brands arguing about which camp is right are losing ground to the ones quietly doing both."
Matthew Berman, founder, Emerald Digital
So what do you actually do?
Run both layers deliberately.
Keep ranking for the searches that matter, because that's how you get retrieved. Then build the citation signals on top: a clear, consistent brand entity, agreement about you across the sources the models trust, coverage of the different ecosystems each engine pulls from, and content rooted in experience only you've. Measure it by citation share and whether you show up when the AI answers a category question, not just by blue-link position.
And keep your honesty intact, because the field is early. Anyone selling settled, precise AI-visibility numbers is overselling. The right move is to start building the signals now, while most of your competitors are still arguing about the acronyms.
The practical move is not to pick a camp. Keep the SEO layer that gets you found, then build the citation layer that makes you worth naming. That is the part most brands are missing.
See how we get brands cited by AI search: we will map where you are being retrieved, where you are being ignored, and what needs to exist before answer engines should name you.
