Last Updated: March 26, 2026
- Google now reports traffic from AI Overviews in Search Console, so impressions, clicks, and position for many queries look different than they did a few years ago.
- You can break out AI Overview performance with dedicated filters, compare it to classic web results, and see how much reach you gain, even when clicks do not keep pace.
- Content that is clear, factual, and structured around real questions tends to surface more in AI summaries, which changes how you plan topics, sections, and page layouts.
- Your reporting needs to shift from only chasing CTR to tracking reach, assisted clicks, and conversions across both AI and traditional search.
Google’s AI Overviews are now part of everyday search, and Search Console quietly reflects that reality in the background, whether you like it or not.
If your organic reports feel different from what you remember, you are not crazy: AI-generated answers have changed how impressions, clicks, and even the meaning of “position” work for a lot of queries.
Where AI Overviews Sit In Search And In Search Console
Google’s old experimental labels like “SGE” or “AI Mode” are gone in practice, and what you see now as a user is usually called “AI Overviews.”
In most markets, these AI answers appear above or mixed in with normal web results and can expand, collapse, and trigger follow-up questions in the same interface.
How Google Describes AI Overviews Today
AI Overviews behave like a summary layer that sits on top of traditional results.
They pull from multiple sources, cite a handful of pages, and then show more web results under the AI block.
“AI Overviews provide synthesized answers for complex or multi-step queries, with citations to web pages that support the response.”
From an SEO and reporting angle, that means the same query can generate impressions in at least two places at once: inside the AI block and in the regular blue links below.
Search Console treats those as related, but not identical, surfaces.
Current Search Console View Of AI Traffic
Search Console now exposes AI-related data in a few ways, and this is where the story actually gets useful.
You are no longer stuck guessing whether AI Overviews are helping or hurting; you can segment traffic and compare patterns over time.
| Area in Search Console | What you see for AI Overviews |
|---|---|
| Search type | Standard “Web” plus, in many properties, a dedicated AI or “AI Overviews” subtype in the Performance report |
| Search appearance | New appearance label for URLs cited in AI Overviews, separate from rich results like FAQ or HowTo |
| Dimensions & filters | Ability to filter or compare queries and pages that appeared within AI Overviews vs all web results |
There are still gaps, of course, and Google will probably keep tweaking these labels.
But you at least have a baseline way to isolate “AI Overview exposure” from the rest of your organic footprint.
How Search Console Counts AI Overview Impressions, Clicks, And Position
The way Google counts impressions and positions around AI Overviews has matured a lot compared to the early experiments.
If you do not understand this part, your reports will keep confusing you, no matter how pretty the dashboards look.
Impressions: When You Actually “Show Up”
Impressions in AI experiences sound simple, but there are a few quirks.
Here is the current pattern that most sites see.
“An impression is recorded when a user can see, or is likely to see, a link or citation to your page in the AI Overview or the surrounding web results.”
- If your page is directly cited inside the AI Overview answer, that counts as an impression for the query.
- If your URL appears in a “web results” list that is part of the AI block, that also counts as an impression.
- Collapsed sections that a user never expands usually do not count until they are made visible on screen.
- Follow-up questions triggered from the same AI experience are treated as separate queries with their own impressions.
This is why informational content often sees a big rise in impressions without a proportional bump in clicks.
You appear a lot more often, but the user might already get enough context in the AI summary.
Position: Where You Stand In Mixed SERPs
Position gets messy when AI Overviews sit above everything else, and Google mixes different units on the same page.
Search Console still compresses this complexity into a single “average position,” but the rules are clearer than they used to be.
| Scenario | How position tends to be reported |
|---|---|
| Your URL is a citation inside the AI Overview tiles | Position reflects the rank of that citation block at the top, usually treated as very high (near position 1) |
| Your URL is in the first organic result under the AI block | Often still counted as position 1 for the organic unit, even though visually it sits below the AI answer |
| Your URL appears both in AI citations and lower organic links | Positions from both surfaces roll into the same average position metric |
This means you can see “average position 1.2” and still feel like you are buried under a large AI box that steals attention.
So position looks strong in Search Console, while your click-through rate tells a very different story.
Quick Reference: How AI Impressions & Positions Are Counted
Here is a compact checklist you can refer back to when you audit reports.
“Treat AI Overview presence as an extra surface, not a replacement for organic, and then read your metrics through that lens.”
- Impressions start when your citation or link can reasonably appear on screen in the AI answer or nearby web results.
- Each follow-up question spun off from the AI interface is a new query with its own impressions and position metrics.
- Average position blends AI citations and classic results, so use Search appearance filters when you want cleaner insight.
- Sudden jumps in impressions for long, question-style queries often mean new AI Overview inclusion, not an algorithm “win.”
This all feels a bit abstract until you line it up with real data.
Once you do, a lot of the “mystery swings” in your charts start to make sense.
What Multi-Year Data Shows About AI Overviews And CTR
At this point, we are not guessing about AI Overviews anymore; we have years of click data, across different industries, to look at.
The patterns are not perfect, but they are clear enough that you can plan around them.
How Organic CTR Has Shifted
Independent studies and internal account data tell a similar story: AI-heavy SERPs change click behavior quite a bit.
Most of the impact shows up on broad, informational queries and on “research” phrases where people are still exploring options.
| Query type | Typical effect from AI Overviews |
|---|---|
| Pure informational (“what is”, “how does”, definitions) | Organic CTR often drops by 20-40%, impressions rise, total clicks are flat or slightly down |
| Research / comparison (“best”, “vs”, “for [use case]”) | CTR usually down 10-25%, but branded sites with strong citations hold up better |
| High-intent commercial (“buy”, “near me”, product names) | AI Overviews show up less often; CTR is closer to historical norms |
| Branded navigational (company or product names) | Minimal change, sometimes slightly higher CTR as brand answers get surfaced more clearly |
If your site is heavy on educational content, you probably felt that first row the hardest.
More people see you, fewer of them click, and the ones who do click often stick longer.
Are AI Overview Clicks “Better” Clicks?
This question sounded hand-wavy a few years back, but by now we actually have some patterns.
The answer is not a simple yes or no, which might sound annoying, but that is how it plays out in the data.
- On several content sites, sessions that start from AI cited queries show higher time on page and more scroll depth.
- On review and comparison sites, AI-initiated visitors click fewer pages per session, but they convert to leads or affiliate clicks at a higher rate.
- On pure news / trending topics, the difference is small; users skim and move on in both cases.
So the typical pattern is: fewer but more serious visits for many informational queries.
If you only watch raw clicks, that feels like a loss; if you watch conversions and assisted conversion paths, it looks a bit different.
Zero-Click Reality You Cannot Ignore
The zero-click conversation used to be overhyped; with AI Overviews, it is finally real in a consistent way.
There are entire query classes where you will rarely “win” a click, no matter how high you rank, because the AI block answers enough on its own.
“If your business model needs a click for every impression, AI Overviews will keep frustrating you; you have to account for pure visibility and citation value too.”
I do not think every brand should accept lower clicks without pushback, but ignoring what users actually see on the SERP is worse.
You need to distinguish between queries you treat as awareness plays and queries where you expect direct traffic and revenue.
How To Break Out AI Overview Traffic In Search Console
The good news is you no longer have to infer AI impact by squinting at line charts.
Search Console gives you enough levers to build a decent AI Overview lens for your reporting.
Step 1: Use Search Type And Appearance Filters
Start in the Performance report and look at the filters along the top.
Depending on your property and region, you may see an AI-related option under search type or under search appearance.
- Open Performance > Search results.
- Set “Search type” to Web to keep the dataset clean.
- Click the “Search appearance” filter and look for any AI Overview or similar label.
- Apply it to see only impressions and clicks where your URLs were part of that AI surface.
If you cannot see a dedicated AI label yet, do not assume Google is not tracking it.
Sometimes labels lag behind rollouts, so you may need to rely more on query segmentation for a while.
Step 2: Build Query Segments For AI-Heavy Searches
Even with appearance filters, segmenting queries is still useful.
Most AI Overviews show up on long-form questions and exploratory searches, not on tight, transactional phrases.
- Add a query filter in Search Console for terms that contain keywords like “how”, “what”, “why”, “best”, “vs”, “for [use case]”.
- Save this as a “Question Queries” segment for quick reuse.
- Compare it against a “Commercial” segment with modifiers like “price”, “buy”, “near me”, “software”, “tool”.
This simple split shows you where AI Overviews are changing your reach and CTR the most.
It also gives you more honest before / after comparisons across rollout periods.
Step 3: Compare Pre And Post AI Behavior
To really see impact, pick a reasonable pre-AI period and a post-AI period and compare the same query groups.
Try not to obsess over single weeks; go for several months on each side if you have the history.
“For core queries, a sustained CTR drop of 30% or more over 8+ weeks, with stable or better average position, is a genuine AI or SERP layout problem, not just noise.”
- For each segment, track impressions, clicks, CTR, and average position.
- Check which landing pages show rising impressions but flat or falling clicks after AI expansion.
- Layer analytics data to see whether engagement improved or worsened on the traffic that remains.
You want a short list of “AI-sensitive” queries and pages that you actively watch each month.
That is much more useful than staring at your entire property graph and wondering why everything feels off.
Step 4: Map What You See To Actual SERPs
Data alone is not enough here, you need to pair it with real SERP checks.
Search your priority queries in an incognito window, on both desktop and mobile, and pay attention to where AI Overviews appear.
- Note which of your pages are cited in the AI block, if any.
- Check whether your classic listing is pushed below the fold on mobile.
- Capture screenshots for internal reporting so you can show what users actually see.
When you line up screenshots with Search Console trend lines, you get a story your team or clients can understand.
Without that, it is very easy for people to misread the numbers and blame the wrong thing.
Shaping Your SEO Strategy Around AI Overviews
Now the harder part: what you actually change in your strategy, instead of just changing how you report.
“Keep doing what works for regular search” is only half true at this point; the details make a big difference.
Make Your Content Easy For AI To Quote
AI Overviews lean toward content that is clear, current, and easy to extract from.
If your key points are buried in long, meandering paragraphs, you are handing advantage to someone else.
- Open each important page with a short, fact-packed answer to the core question in 2-3 sentences.
- Use descriptive subheadings that match the way people actually search, including full questions.
- Break complex steps into ordered lists or short paragraphs that read cleanly out loud.
- Refresh stats, prices, and time-sensitive claims regularly so you stay competitive with fresher sources.
This is not about gaming the system; it is about being the easiest credible source for Google’s models to interpret.
If that sounds boring, good, because most real SEO wins are boring and consistent.
Double Down On E‑E‑A‑T Signals
AI-generated summaries lean heavily on sites that look real, trustworthy, and accountable.
If your site feels anonymous or thin on ownership, you are handicapping yourself.
- Add clear author bylines with real names and credentials for any content that gives advice or analysis.
- Build out detailed About and Contact pages, including business registration details where relevant.
- Link to credible external sources and references, especially in YMYL topics like health, money, or legal areas.
- Use Organization and Person schema so Google can tie your content back to a real entity.
I know some marketers still treat E‑E‑A‑T like a fuzzy concept, but in practice it is just structured credibility.
Without these signals, your content can rank for random long-tail terms but still struggle to be cited in AI Overviews.
Structured Data: No “AI Schema,” Still Very Useful
Google has not shipped a magical “AI Overview schema” flag, and I would not wait for one.
What it has done is keep rewarding clean, consistent structured data that explains what your content covers.
- Make sure FAQ, HowTo, Product, Review, and Article schema are valid on pages where they apply.
- Use ClaimReview for fact checks or myth-busting content where you evaluate specific claims.
- Include clear dates, authors, and publisher information in schema for news and educational posts.
- Keep schema in sync with the visible page; mismatches just reduce trust.
The goal is to make it trivial for Google’s systems to understand your page type, topic, and expertise.
Once that is solid, you have a much better shot at being surfaced in both AI Overviews and regular rich results.
Design Content For Follow-Up Questions
AI Overviews invite users to ask follow-up questions in the same flow, and those follow-ups often drill deeper on the same topic.
If your content stops at the first layer, you miss those extra queries and the stronger intent behind them.
- Map 3-5 likely follow-up questions for each core topic and address them clearly in sub-sections.
- Use internal links to spin those follow-ups into their own focused pages when the topic is big enough.
- Watch your query reports for new long-tail phrases that start appearing after AI expansion and build content around them.
Think in paths, not just single keywords.
The queries that show up two or three steps into an AI conversation can be where the best-qualified visitors come from.
Controls: When To Limit Snippets Or Indexing
There are cases where you might not want your content summarized or heavily quoted inside AI Overviews.
That is not an emotional decision; it is usually about risk, context, or business model.
“Use nosnippet, max-snippet, and noindex as levers, not as panic buttons.”
- Medical, legal, or financial guidance: Consider stricter snippet limits if misquotes could cause harm or liability.
- Membership or paywalled content: You might keep summaries shallow to avoid leaking too much value outside the paywall.
- Premium research or proprietary data: Use shorter max-snippet values so AI Overviews can mention you, but not recreate your full findings.
- Low-value or temporary pages: Noindex them if you do not want them appearing in any search surface, AI or otherwise.
The trade-off is simple: more exposure vs more control.
For most public content, I would lean toward exposure, but high-risk verticals should be more cautious and deliberate.
Rethinking Reporting And Communication Around AI Overviews
This is where a lot of SEO work lives now: not only doing the right things, but explaining what the numbers actually mean in this new context.
If you report the old way, you set the wrong expectations and everyone feels like they are losing, even when reach and revenue are stable.
Build A Simple Reach-Engagement-Value Story
I like to reduce search performance into three layers that anyone on the team can understand.
It turns AI Overviews from a scary black box into one part of a clear narrative.
- Reach: Impressions from both classic web results and AI Overview appearances.
- Engagement: Clicks, CTR, time on page, pages per session, and scroll depth.
- Value: Conversions, leads, or revenue attributed to organic, including assisted conversions.
Your job is to show where AI Overviews expand reach, where they compress engagement, and where value holds steady or improves.
Once stakeholders see that full arc, isolated CTR drops do not instantly trigger panic.
Talking About AI Overviews With Clients Or Internal Teams
Many non-SEOs still have a fuzzy understanding of what AI Overviews do to search results.
If you jump straight into graphs, you lose them; start with what users see on the page.
- Explain that Google now answers many questions with AI summaries that cite a few sites and then list more results below.
- Clarify that Search Console includes those impressions and clicks, and that this changes metrics for some queries more than others.
- Set the expectation that informational queries might show higher reach but lower CTR, while commercial and branded searches behave more like they used to.
- Show at least one real SERP screenshot next to the performance chart for that query group.
People trust numbers more when they can tie them back to something they recognize on their own screen.
Without that, it is easy for AI conversations to turn into vague fear or unhelpful debates.
Where To Focus Your Effort Next
If you try to react to every search change, you burn out and your strategy frays.
I would focus on a tight list of moves that keep paying off as AI Overviews keep evolving.
- Harden your technical base: fast, mobile-friendly pages, clean internal linking, and valid structured data.
- Invest in clear, expert content with strong introductions and direct answers that AI can easily quote.
- Track a set of high-impact query segments with AI exposure and review them on a fixed schedule.
- Use analytics to judge traffic quality, not just volume, especially for your top informational assets.
Google will keep adjusting how AI Overviews look and how they show up in reports, and sometimes that will be annoying.
But if you keep your measurement sharp and your content genuinely helpful, you stay in a good place, even when the presentation layer keeps changing.
Need a quick summary of this article? Choose your favorite AI tool below:
