On June 3, 2026, Google announced dedicated Search Console reports for website visibility in generative AI features. The Search report covers AI Overviews and AI Mode, while a separate report covers generative AI experiences in Discover. For the first time, participating site owners can examine this visibility without trying to infer it from screenshots, third-party estimates, or manual searches.
The release matters to small businesses because search behaviour is changing while commercial questions remain familiar. A Toronto business owner may ask Google to compare website providers, explain a Microsoft 365 problem, or recommend considerations for managed IT. Being visible during that research can help discovery, but visibility alone does not show whether the person visited, contacted, or hired the company.
What the generative AI performance report shows
The report shows how often links to the site appeared in supported generative AI features on Google Search. Google currently lists AI Overviews and AI Mode. Search Labs experiments are not included, and the supported feature list can change as Search develops.
| Dimension | What it answers | Useful business question |
|---|---|---|
| Impressions | How often a site link appeared in a supported AI feature | Is visibility growing, stable, or concentrated in short spikes? |
| Pages | Which final canonical URLs received impressions | Which service pages, guides, or resources are being surfaced? |
| Countries | Where the searches originated | Does visibility match the markets the business can actually serve? |
| Devices | Whether visibility occurred on desktop, tablet, or mobile | Does the landing experience work well on the devices customers use? |
| Dates | How impressions changed by day, week, or month | Did a sustained change follow publication, an update, seasonality, or a broader search shift? |
Most performance data is assigned to the canonical URL. The chart is aggregated at property level unless a URL filter is applied, while page rows are aggregated by URL. This matters because chart totals and table totals can differ without either being wrong. Google also applies the usual Search Console limitations, including table row limits and preliminary values for recent data.
What an AI search impression means, and what it does not
An impression means a link to the site was shown to a user in a supported generative AI feature. It is evidence of presentation, not proof that the person noticed the link, agreed with the surrounding response, clicked through, became a qualified lead, or received a direct recommendation for the business.
- Do not present AI impressions as visits or leads.
- Do not convert impression growth into an estimated revenue figure without observed customer data.
- Do not assume one visible page means the entire domain has authority for the subject.
- Do not promise that content or structured data will secure inclusion in an AI response.
- Do compare visibility with clicks, landing-page engagement, meaningful actions, lead quality, and customer outcomes.
1. Build a trustworthy baseline
Begin with a period long enough to reduce daily noise. For many small sites, twenty-eight days compared with the previous twenty-eight days is a practical operating view. Use a three-month or year-over-year comparison when seasonality is important. Record the selected property, date range, filters, and export date so the next report measures the same thing.
- Confirm the preferred domain property and ownership in Search Console.
- Export the AI report before applying interpretation or commentary.
- Record total impressions and the highest-visibility canonical pages.
- Separate Canada from other countries when the business primarily serves Canadian customers.
- Note mobile and desktop distribution before prioritizing page-experience work.
- Label major launches, migrations, outages, campaigns, and tracking changes on the reporting timeline.
2. Group pages by customer intent
A list of URLs becomes useful when the pages are grouped by the job they perform. Separate service pages, educational guides, comparisons, pricing information, case evidence, support resources, and company pages. A practical guide may create broad visibility near the beginning of research, while a service or pricing page is closer to a buying decision. Both can be valuable, but they should not be judged by the same expectation.
| Page role | Likely customer need | Useful next step |
|---|---|---|
| Educational guide | Understand a problem or process | Link naturally to the related service and a deeper resource |
| Comparison | Evaluate approaches or providers | Clarify decision criteria, tradeoffs, proof, and scope |
| Service page | Find capable help | Make outcomes, process, fit, evidence, and contact options clear |
| Pricing page | Assess affordability and scope | Explain inclusions, variables, assumptions, and how to request a quote |
| Case evidence | Reduce perceived risk | Show the problem, work, measurable result, and relevant limitations |
If one article receives most AI impressions, inspect whether it introduces the organization clearly and provides a helpful route to the relevant service. Do not turn every educational paragraph into a sales pitch. One contextual internal link and a specific next step are often more useful than repeated promotional interruptions.
3. Compare AI visibility with standard search performance
Google says AI feature data remains included in the overall Performance report under the Web search type. The standard report provides clicks, impressions, click-through rate, and average position across search results. The dedicated AI report isolates supported AI impressions, but it should be read beside overall organic performance rather than replacing it.
- Compare the same dates in both reports.
- Identify pages that gain AI visibility and overall clicks together.
- Flag pages with growing AI impressions but weak overall visits for a landing-page and intent review.
- Watch for overall click changes without assuming that the AI report alone explains them.
- Use weekly or monthly trends for decisions instead of reacting to one day's preliminary data.
4. Connect search visibility to meaningful website actions
Search Console explains what happened before a visit. Website analytics explains what visitors did after arriving. In Google Analytics, a key event is an action important to the business, such as a completed contact form, booked consultation, qualified phone call, quote request, account registration, or purchase. Choose actions that reflect real progress rather than counting every button click as success.
- Track successful form submissions, not only form starts.
- Use click-to-call and WhatsApp events as directional signals, then confirm whether a genuine conversation occurred.
- Record the landing page and relevant service interest in the lead intake process.
- Distinguish spam, job inquiries, vendor messages, and unsuitable requests from qualified opportunities.
- Connect closed work to the original lead where consent, privacy, and system accuracy allow it.
A practical small-business example
Consider an Ontario company with a detailed guide about choosing a website provider and a service page for small-business website design. The guide begins receiving AI impressions, especially from Canada and mostly on mobile. The team checks the mobile layout, clarifies who the advice is for, adds a visible link to the service page, and ensures the consultation form works without unnecessary fields.
The following month, the guide has more AI impressions and the website has more organic form submissions. That is promising, but it is not yet proof of cause. The team reviews landing pages, source data, submission quality, and sales notes. If several qualified Toronto inquiries mention the guide, the evidence becomes stronger. If inquiries remain unsuitable, the better action may be refining service scope and qualification rather than publishing more content.
5. Use a monthly scorecard that supports decisions
| Measure | Why it matters | Decision |
|---|---|---|
| AI feature impressions | Shows visibility in supported experiences | Identify trends and pages worth reviewing |
| Canadian impression share | Checks geographic alignment | Strengthen genuinely relevant Canadian examples or service information |
| Organic clicks | Shows visits from Google Search | Review title, intent, competition, and page usefulness |
| Organic key events | Shows meaningful on-site actions | Improve the journey and calls to action |
| Qualified leads | Separates activity from commercial fit | Refine messaging, offer, and intake |
| Won work or revenue | Connects marketing to business outcomes | Invest in the topics and services that create sustainable value |
Add a short interpretation below the numbers: what changed, the most likely explanations, what evidence is missing, and the next action. Avoid attributing every increase to the latest content edit. Search demand, news, seasonality, competitors, brand activity, and reporting changes can all affect the trend.
6. Improve eligibility and usefulness with established SEO practices
Google states that there are no additional technical requirements to appear as a supporting link in AI Overviews or AI Mode. A page must be indexed and eligible to appear in Google Search with a snippet. Google also says no special AI schema, AI text file, or new machine-readable markup is required for these features.
- Make important information available in crawlable, accessible HTML.
- Answer the customer's real question clearly and support it with original experience, examples, comparisons, or evidence.
- Use descriptive titles, logical headings, canonical URLs, and helpful internal links.
- Keep structured data accurate and consistent with visible content.
- Provide relevant high-quality images and video where they improve understanding.
- Maintain a fast, secure, mobile-friendly page experience.
- Keep company, service, location, and Google Business Profile information current.
Useful content is not a volume contest. Review the pages already earning impressions, improve outdated or incomplete material, and publish a new page only when it answers a distinct customer need. Original project evidence, clear authorship, current sources, and honest limitations are more durable than speculative tactics designed for one interface.
A practical 30-day implementation plan
| Week | Work | Output |
|---|---|---|
| Week 1 | Verify Search Console, analytics, key events, consent, and lead definitions | Documented measurement map and current baseline |
| Week 2 | Export AI and standard performance data, then group priority pages by intent | Page-level visibility and outcome scorecard |
| Week 3 | Review the highest-opportunity page for usefulness, internal links, mobile experience, and conversion path | One controlled page improvement with a dated annotation |
| Week 4 | Compare early signals, inspect qualified inquiries, and choose the next test | Monthly report with evidence, limitations, and one accountable action |
Common measurement mistakes to avoid
- Searching manually and treating one personalized result as a market-wide ranking report.
- Calling every AI impression a citation, recommendation, visit, or lead.
- Comparing date ranges of different lengths or ignoring normal seasonal demand.
- Changing several pages, offers, and tracking rules at once, then claiming one caused the result.
- Publishing near-duplicate pages for every city or question to manufacture visibility.
- Using third-party visibility estimates as if they were first-party Search Console data.
- Reporting traffic without checking whether inquiries match the services and regions the business serves.
Monors helps Toronto and Ontario businesses improve the technical, content, measurement, and conversion foundations behind organic search. A review can connect Search Console, website analytics, service pages, mobile performance, structured data, and lead handling into one practical plan. The goal is not to promise an AI recommendation; it is to make useful visibility measurable and every customer path easier to understand.