AI Reports in Google Search Console: How to Read Them

AI report in Google Search Console helps you track how prominently your website appears in Google’s two generative search experiences: AI Overviews and AI Mode. However, this is primarily a visibility report, not a complete traffic dashboard. You should use impressions to identify which content Google selects as a source, then compare them with Performance, Google Analytics, and conversion data to evaluate business value.

Google began introducing the Search Generative AI performance report on June 3, 2026. The report is being rolled out in phases and may not yet be available for every property. The report currently includes impressions from AI Overviews and AI Mode; experiments in Search Labs are not included (according to Google Search Console Help).

What does the AI report in Google Search Console include?

You can usually find the report in the Performance section or under the performance area related to Generative AI, depending on the interface and your property access. If you cannot see it yet, Google may not have enabled the feature for your property, or the website may not have enough impressions in AI features.

The report focuses on four main analysis groups:

Impressions: the number of times links to your URLs were displayed in an AI feature on Google Search.
Pages: the URLs Google included in AI Overviews or AI Mode.
Countries: the countries where searches originated.
Devices: search devices, including computers, tablets, and mobile devices.

You can also view trends by day, week, or month. Daily data uses Pacific Time, so the report may show different dates from your internal analytics systems in Vietnam. The latest data may be preliminary and can change during the first few hours (according to Google Search Central).

The important point is that the AI report does not replace the standard Search report. Google says AI data is still included in the overall search performance overview, while the Generative AI report provides a separate view that isolates visibility in generative features.

How to read AI Overviews impressions and AI Mode SEO data

1. Read total impressions over time

Start by selecting a sufficiently long period, such as 28 or 90 days. A single spike is not enough to conclude that your content is being prioritized. AI Overviews and AI Mode can change their display scope, interface design, and source-selection methods.

Instead of asking only, “Are impressions increasing or decreasing?”, ask three questions:

Has the upward trend continued for multiple weeks?
Is the increase coming from a few major pages or being distributed across many URLs?
Are impressions increasing in a particular country, device, or content group?

2. Analyze by Pages

The Pages table is the most valuable area for content strategy. If an article has many AI Overviews impressions, this indicates that Google is using or linking to the content in a generative answer. This is a signal of topical relevance, but it does not mean the article is certain to receive many clicks.

Group URLs by purpose:

Informational pages: definitions, guides, and processes.
Comparison pages: help users evaluate options.
Commercial pages: products, services, and pricing pages.
In-depth pages: research, analysis, and technical documentation.

For example, if guide articles have high impressions but service pages barely appear, you can build internal links from informational articles to conversion pages, add evidence, scope of application, frequently asked questions, and clear calls to action.

To strengthen your content foundation, you can also see the guide to optimizing blogs for AI Overviews and AI Mode with principles covering structure, citability, and reading experience.

3. Compare AI Overviews with AI Mode

These two features should not be treated as a single concept. AI Overviews typically appear as a summarized layer on the results page, while AI Mode focuses more on a conversational experience and follow-up queries. A URL may be a good fit for AI Overviews because it answers the initial question well, but it may not necessarily be used in expanded queries in AI Mode.

Criterion	AI Overviews	AI Mode
Role	Summarizes an answer on the results page	Conversational search experience with multiple turns
Type of need	Questions requiring a quick synthesis	Exploration, deeper investigation, and question refinement
Data in the AI report	Impressions of displayed links	Impressions of displayed links
Metrics to compare	Search impressions, clicks, CTR	Trends by page, country, device, and overall traffic
Appropriate SEO actions	Direct answers with clear structure	Add depth, context, and related questions
Limitations	Clicks cannot be inferred from impressions alone	It is difficult to accurately identify the query session that led to a visit

For AI Mode SEO, prioritize content capable of answering multiple layers of questions: applicable conditions, exceptions, how to implement, costs, risks, and next steps. Do not just write a short definition and expect the content to be used throughout the search journey.

How to measure traffic from AI Search more accurately

At present, the Generative AI report in Search Console primarily provides impressions. Therefore, you should not treat this figure as actual traffic. The measurement process should include three layers of data.

Layer 1: Measure visibility

Export data by Pages, Countries, Devices, and Dates. Create a tracking table containing the URL, AI impressions, total Search impressions, Search clicks, CTR, country, and content group. Track this at least weekly to reduce the impact of short-term fluctuations.

Layer 2: Compare visits

In the traditional Search Console, view the same period by Page and Query. Compare three cases:

AI impressions increase and Search clicks increase: a positive signal, but it does not prove that all clicks came from AI.
AI impressions increase and Search clicks remain unchanged: the content may be visible, but the link may not be compelling enough, or users may have already received the answer directly on Google.
AI impressions decrease and Search clicks increase: AI is not the URL’s primary source of visibility; do not rush to revise the article if overall performance remains strong.

In Google Analytics or another analytics tool, check the landing page, traffic source, engaged sessions, sign-ups, and conversions. Google Search referrers generally do not clearly distinguish users arriving from traditional results, AI Overviews, or AI Mode, so interpret the data cautiously rather than attributing all organic traffic to AI Search.

If your goal is conversion, add campaign parameters to links you control, such as newsletter links, sharing buttons, or remarketing campaigns. You cannot add UTM parameters yourself to links that Google displays in AI Overviews or AI Mode.

Layer 3: Measure business quality

Impressions only have meaning when tied to a goal. A page with few impressions that generates sign-ups, calls, or revenue may still be more important than an article that receives many impressions but prompts no action.

You should also track:

Conversion rate by landing page.
Number of leads from content that appears in AI.
Revenue or order value by page.
Returning-user rate and email sign-ups.

Limitations, common errors, and action process

Common reporting errors

Confusing impressions with clicks: a displayed URL does not mean that the user visited it.
Comparing different time zones: Search Console uses Pacific Time, while Analytics may use the account’s time zone.
Treating the latest data as final: preliminary data may change.
Comparing the chart total with the table without considering how data is aggregated: the chart may aggregate by property, while the Pages table aggregates by URL.
Drawing conclusions from a single URL: you need to examine topic groups and a sufficiently long period.

Weekly review process

Export the AI report for the most recent 7 or 28 days.
List the 10 URLs with the largest increases in impressions and the 10 URLs with the largest decreases.
Compare each URL with Search clicks, CTR, and conversions.
Check the country, device, and canonical URL to rule out technical discrepancies.
Choose no more than three actions: improve the direct answer, add supporting evidence, increase internal links, or fix the conversion experience.
Reassess after at least several weeks; do not make wholesale changes based on a single short-term fluctuation.

If your website has a lot of content that needs updating, also see AI-era SEO strategy for content. If you want to turn promising articles into multiple touchpoints, the guide to repurposing blog content can help expand distribution beyond Google.

In summary, the AI report in Google Search Console should be viewed as a data layer showing visibility in AI Overviews and AI Mode. The correct approach is to use impressions to identify opportunities, use Search Console and Analytics to verify traffic, and then use conversions to determine priorities. Do not optimize merely to increase impression counts; focus on pages that are both selected by AI as sources and helpful in guiding readers to take the next step.

What Does Google Search Console’s AI Report Actually Measure?