Module 02 · Lesson 1 of 4
Google AI Overviews and AI Mode
Google AI Overviews are generated summaries shown among traditional results for some queries. AI Mode is a full conversational search experience where the exchange is question-and-answer rather than ten blue links. Both generate text and both display links to sources alongside that text. The surfaces change often; the way to learn them is observation, not memorising screenshots.
What Google actually says
Three statements from Google's documentation carry most of the weight for this whole module. One: the AI features draw on Google's core search systems, so the fundamentals you learned in Module 1 decide who is even in the pool. Two: there are no special schema types or additional technical requirements to be eligible; ordinary, well-built pages are the baseline. Three: being eligible does not guarantee inclusion in any generated answer. Eligibility is entry to the competition, not a ticket to the podium.
How answers present sources
| Surface | What the user sees | Where your visibility comes from |
|---|---|---|
| Classic results | Ranked links with titles and snippets | Ranking: relevance, authority, experience |
| AI Overviews | Generated summary with links to cited sources | Being selected as a source for the generated answer |
| AI Mode | Conversational answers with follow-ups and cited links | Source selection across the conversation, including follow-up prompts |
Notice the shift in the third column. Classic search ranks pages against each other. AI surfaces assemble answers from selected sources. Ranking still matters (the pool is drawn from indexed, credible content), but a new question sits on top of it: is this page the kind of source an answer gets built from?
Interfaces move; method does not
Button labels, placements and availability of these surfaces change quarter to quarter and country to country. Do not build strategy on pixel positions. Build it on the stable layer underneath: indexed, clear, credible pages that answer questions in extractable form, corroborated by sources the systems already trust. Every later module works that layer.
- Take three queries from different intents: a how-to ("how to fix crawl errors"), a comparison ("semrush vs ahrefs"), and a definitional one ("what is e-e-a-t").
- Search each in Google and record: AI Overview present or not, AI Mode available or not, which domains the generated answer credits.
- Repeat the definitional query with one extra layer of specificity ("what is e-e-a-t for local businesses") and note how the credited sources change.
Run it yourself this week; your observations are the lesson. The exact results you see matter less than the habit of logging them.
- Choose five queries from your Workbench 1.2 list, one per intent type where possible.
- For each, log in a sheet: date, surface present (AI Overview / AI Mode / neither), and the first three cited domains if an answer appeared.
- Keep the sheet. It becomes the seed of the visibility log you will build properly in Module 3.
Does being eligible guarantee appearing in an AI Overview?
Do you need special schema to be eligible for Google's AI features?
What is the practical difference between AI Overviews and AI Mode?
Sources used in this lesson
Google Search Central: AI features and your website