Does AI search reward the same signals as classic ranking, or different ones?
- Published November 6, 2026
- Linked from 1 answer
- In a cluster of 63
In Google Search, the same foundation applies, by Google’s own description. Its AI features documentation sets no extra requirements for AI Overviews or AI Mode beyond being indexed and eligible to appear with a snippet, and no special optimizations; its general SEO best practices are what it points to. For AI search tools outside Google, such as ChatGPT search and Perplexity, the companies publish how their crawlers work but not how they rank or choose sources, so claims about their “signals” are guesses.
What is shared and what differs in Google
| Classic results | AI Overviews and AI Mode | |
|---|---|---|
| Eligibility | Indexed pages | Indexed pages eligible for a snippet |
| Controls | <!–INLINECODE0–>, <!–INLINECODE1–>, <!–INLINECODE2–>, <!–INLINECODE3–> | The same controls, plus a Search generative AI control in Search Console that excludes a site from these features only |
| How a response is assembled | Ranked list of results for the query | A generated answer with supporting links, which may draw on several related searches (query fan-out) |
| Where traffic is reported | Search Console Performance report, Web search type | Clicks in the same report and search type; impressions also in a separate Generative AI performance report |
The third row is the real difference. Because an AI response can draw on several related searches, a page can be a supporting link for a question it does not rank first for, and a page that ranks first for the main query can be left out. That changes which pages appear, not what makes a page good.
What that means for the work
The signals Google describes for ranking, such as relevance, helpful and reliable content, links from other sites, and page experience, are not replaced by a separate AI set. Google’s guidance on creating helpful content applies to both. The adjustment is editorial: pages that state clear answers to the questions around a topic, not only to its headline query, give a multi-search answer more to draw on.
Outside Google
OpenAI and Perplexity each document a crawler for their search features, OAI-SearchBot and PerplexityBot, and both say these are separate from the crawlers used for model training. Allowing them is the documented requirement for appearing; how either tool chooses among eligible pages is not public. For the steps that apply to those tools, follow how to get ChatGPT or Perplexity to cite your page. Treat any detailed list of AI ranking factors for them as someone’s observations, and test it against your own results before acting on it.
Add a comment