Google AI Mode, AI Overviews and organic search results can all help people explore a topic, but they are different surfaces. A dataset becomes confusing when it uses the word “ranking” for every kind of appearance or treats one AI experience as a substitute for another.
For an SEO or research workflow, decide which experience you are observing before defining your metrics. Here is how the distinction applies to ProtoScrape’s current API and the reports you can build from it.
Google AI Mode is an AI search experience
Google describes AI Mode as useful for nuanced questions, comparisons and exploration, with responses that include links to supporting websites. A useful observation can therefore include answer text and available sources rather than only an ordered list of pages.
ProtoScrape’s currently available google_ai_mode engine returns structured answer text and available citations for a submitted query. It is suited to questions such as “Does this answer mention a product?” and “Which pages were returned as citations?” It is an independent service, not Google’s own API.
AI Overviews appear within search results
AI Overviews are summaries that may appear in Google Search when Google determines that they add value to the results. They are not present for every query. Google also explains that AI Mode and AI Overviews may use different models and techniques, so their answers and links can differ.
Do not label an AI Mode observation as an AI Overview. A citation returned in one experience is not evidence that the same page appeared in the other. ProtoScrape does not currently expose a separate AI Overviews collection endpoint.
These distinctions follow Google’s explanation of its AI search features, which is the primary reference for how Google describes the experiences.
A SERP observation answers a different question
SERP means search engine results page. For an organic-result dataset, useful fields include the ordered result position, page title, URL and snippet. That supports questions about which pages appeared in the observed result list for a query.
A source listed in an AI answer should not be assigned an organic ranking merely because it appears first in a citation array. Likewise, a high organic position does not establish that the page was cited in an AI answer. Store the surface and metric definitions explicitly when your reports combine different types of observations.
Match the dataset to the research goal
| Research question | Observation to collect |
|---|---|
| Does an AI Mode answer name our product? | AI Mode answer text |
| Which domains appear as AI Mode sources? | AI Mode citation URLs |
| Which pages appear in an organic result list? | Organic SERP results |
| Is an AI Overview shown, and what does it say? | A separate AI Overview observation |
The table describes the data needed for each question. It is not a list of features currently available from ProtoScrape. Today, Google AI Mode is in early access; Google SERP, ChatGPT and Perplexity remain in validation. Check the engine availability endpoint before scheduling work.
Keep comparisons honest
Use the same question set and supported market settings where possible, record collection time, and retain failed-observation counts. Separate brand mentions, source citations and organic positions in the report. Repeated collection is useful for studying changes, but it cannot guarantee what another person will see.
If AI Mode is the surface you need, start with the API integration guide. For measurement definitions and sample-size reporting, continue to AI visibility tracking.