Most AI search reporting still lives in five different places: a GA4 exploration someone built once, a Search Console tab nobody exports, a prompt-tracking tool with its own login and a spreadsheet of manual ChatGPT checks. When the leadership team asks "are we showing up in AI answers, and is it worth anything?", the answer takes a week to assemble.
A single dashboard fixes that. In this guide we walk through building an AI search dashboard in Google's free reporting tool, combining three data sources: GA4 AI referrals, Search Console's new generative AI data and exports from a prompt-tracking tool. One naming note first: in April 2026 Google renamed Looker Studio back to Data Studio. The product, connectors and workflow are the same, so if you know Looker Studio, you already know how to do this.
You will learn which fields to pull, how to blend them without double counting, which charts to build and the mistakes that make AI dashboards misleading.
Key Takeaways
- Looker Studio is now called Data Studio (renamed 16 April 2026); the free GA4 and Search Console connectors still work the same way.
- GA4 now has a built-in AI Assistant default channel (added 13 May 2026) for visits from tools like ChatGPT, Gemini, Copilot and Claude.
- Clicks from Google AI Overviews and AI Mode are counted in Organic Search, not the AI Assistant channel, so you cannot isolate them in GA4.
- Search Console's Generative AI performance report shows impressions only by page, country, date and device; export it to a Google Sheet to use it in a dashboard.
- Prompt-tracking data (mentions, citations, share of voice) comes from third-party tools via CSV or Sheets, and should be labelled as sampled data.
What Changed in 2026 (and Why Your Old Dashboard Is Out of Date)
Three platform changes in 2026 make it worth rebuilding any AI dashboard you set up last year.
- The Data Studio rebrand. Google's release notes for 16 April 2026 say it plainly: "We've rebranded Looker Studio as Data Studio." Gemini in Looker was renamed Gemini in Data Studio at the same time.
- GA4's AI Assistant channel. Until May 2026 you had to build a custom channel group with a regex for chatgpt.com, perplexity.ai and similar referrers (the method we described in our guide to tracking ChatGPT and AI traffic in GA4). GA4 now does this by default.
- Search Console's Generative AI performance report. After a test that started in June 2026, Google confirmed it had rolled the report out to all websites worldwide as of 31 August 2026.
“Google Analytics now provides a dedicated way to measure and analyze traffic originating from popular AI assistants.”
— Google Analytics Help, What's new (13 May 2026), What's new in Google Analytics
The catch is that each source measures something different. GA4 counts sessions that reached your site. Search Console counts impressions inside Google's AI features. Prompt trackers count how often you are mentioned or cited in a sample of prompts. A good dashboard keeps these separate and explains each one, rather than adding them together.
The Three Data Sources and What Each Can Tell You
| Source | How it connects | What it measures | Main limitation |
|---|---|---|---|
| GA4 (AI Assistant channel) | Native Google Analytics connector | Sessions, engaged sessions, key events and revenue from ChatGPT, Gemini, Copilot, Claude and similar assistants | Excludes AI Overviews and AI Mode; misses visits with no referrer |
| Search Console (standard Performance) | Native Search Console connector (Site or URL Impression table) | Clicks and impressions for Web search, which include AI Overviews and AI Mode | Cannot separate AI feature clicks from classic results |
| Search Console (Generative AI report) | Export to Google Sheets, then Sheets connector | Impressions in AI Overviews and AI Mode by page, country, date and device | No clicks, no queries; 1,000-row table limit |
| Bing Webmaster Tools AI Performance | Export to spreadsheet | Citations and grounding queries in Copilot and Bing AI summaries | Grounding query data is a sample |
| Prompt-tracking tool | CSV/Sheets export or a partner connector | Mentions, citations, share of voice, sentiment across chosen prompts | Sampled; results vary run to run |
If you are new to the GA4 side, our guide to GA4 referral traffic explains where AI referrers used to sit before the new channel existed.
Step-by-Step: Build the Dashboard
- Create a GA4 data source. In Data Studio, choose Create > Data source, pick the Google Analytics connector and select your GA4 property. Use the dimension Session default channel group and filter to "AI Assistant". Add Session source, Landing page + query string, sessions, engaged sessions, key events and (for ecommerce) purchase revenue.
- Create two Search Console data sources. Google's documentation explains that the connector offers a Site Impression table and a URL Impression table and that a single data source can only use one, so add both if you want site totals and page-level detail in one report.
- Export the Generative AI report weekly. In Search Console, open the Generative AI performance report, set the date range and export the Pages table. Paste it into a Google Sheet tab with a "week" column, then connect the Sheet. Google's help notes that values shown as "~" or "-" export as zeros, so treat zeros with care.
- Add Bing AI citations. Bing's AI Performance report can now be exported to a spreadsheet. Append it to a second Sheet tab with the same week column.
- Add prompt-tracking exports. Export weekly mentions, citations and share of voice per prompt cluster. Keep the prompt set fixed; our guide to prompt research covers how to choose one.
- Blend carefully. Only blend sources that share a real join key, such as landing page URL (GA4 landing page and Search Console page) or week. Normalise URLs first: strip query strings, force lowercase and use the canonical version.
- Build four pages. An executive summary, a Google AI page, an assistant referrals page and a prompt visibility page (layout below).
- Annotate. Add a text box with data notes: definitions, known gaps and the date each source was last refreshed.
Recommended Dashboard Layout
| Page | Charts | Source | Question it answers |
|---|---|---|---|
| 1. Summary | Scorecards with month-on-month change; one trend line per source | All | Are we more visible and is it driving business? |
| 2. Google AI features | AI impressions by page (table), by country (geo map), trend (time series) | Generative AI report export | Which pages Google shows in AI Overviews and AI Mode |
| 3. AI assistant referrals | Sessions by source, landing page table with engagement rate and key events | GA4 | Which assistants send visitors and do they convert |
| 4. Prompt visibility | Share of voice by competitor, citation rate by prompt cluster, sentiment | Prompt tracker, Bing export | How often we are mentioned versus competitors |
Keep the summary page to six scorecards at most. If you want ideas for scorecard design, our older guide to building an SEO dashboard in Looker Studio still applies, and the KPI definitions are covered in our companion post on AI search KPIs to report each month.
Useful calculated fields
- AI referral conversion rate:
SUM(Key events) / SUM(Sessions)filtered to the AI Assistant channel. - AI share of referrals: AI Assistant sessions divided by all sessions, to show the scale honestly (it is often small).
- Page group: a
CASEstatement on landing page path (for example blog, service, product) so you can see which content types AI favours. - Clean URL:
REGEXP_REPLACE(Landing page, "\\?.*$", "")to strip parameters such asutm_source=chatgpt.combefore joining.
How to Read the Data Honestly
The biggest risk with AI dashboards is false precision. A few rules keep the numbers honest:
- Never add Search Console AI impressions to Web impressions. They are already included in the Web search type totals.
- Do not label Organic Search as "non-AI". Google's channel documentation says Organic Search includes Google's AI Overviews and AI Mode.
- Show dark traffic as a caveat. Some AI visits arrive without a referrer and land in Direct; see why AI traffic shows as direct.
- Use ranges and trends for sampled data. Prompt-tracking results change between runs, so report four-week averages rather than single days.
For a deeper look at the Google side, our post on Search Console AI data explains the Generative AI report's dimensions and limits in detail, and AI share of voice covers how to calculate the competitor view on page four.
Common Mistakes
- Treating the AI Assistant channel as "all AI traffic". It excludes Google's AI features and any visit without a referrer.
- Blending on mismatched keys. Joining GA4 landing pages that include query strings to Search Console's canonical URLs produces empty or duplicated rows.
- Assuming the Generative AI report is in the connector. At the time of writing, Google's Data Studio connector documentation lists Web, image, video, news and Discover search types but does not mention generative AI data, so plan on exports.
- Letting exports lapse. If nobody exports the Generative AI report weekly, you lose history. Put it on a calendar or automate it.
- Changing the prompt set every month. Trends become meaningless. Add new prompts as a separate cluster instead.
- Reporting vanity numbers without outcomes. Pair visibility with key events and revenue on the same page.
Related Guides
- How to Choose an AI Visibility Tracking Tool: Questions to Ask
- AI Citation Audit: Find Which of Your Pages AI Already Cites
- AI Search KPIs: The Metrics to Report Each Month
Frequently Asked Questions
Is Looker Studio still available in 2026?
Yes, under a new name. Google rebranded Looker Studio as Data Studio in April 2026. Existing reports, data sources and connectors continue to work, so you can follow older Looker Studio tutorials.
Can I see AI Overview clicks in GA4?
Not separately. Google's GA4 channel documentation says Organic Search includes AI Overviews and AI Mode, and the new AI Assistant channel excludes them. Use Search Console's Generative AI report for AI feature impressions instead.
Does the Search Console connector include the Generative AI performance report?
The Data Studio connector documentation does not mention it at the time of writing. Export the report from Search Console to Google Sheets and connect the Sheet, and check Google's documentation for updates.
How often should I refresh an AI search dashboard?
GA4 and Search Console connectors refresh automatically. Export the Generative AI report, Bing AI Performance and prompt-tracking data weekly, and review the dashboard monthly with stakeholders.
Which prompt-tracking tools connect to Data Studio?
Many tools offer CSV or Google Sheets exports, and some have partner connectors. Check each vendor's documentation, and see our buying guide for questions to ask before you choose.
Turn Your AI Data Into Decisions
A good AI search dashboard is not about one magic number. It shows where Google features your pages, which assistants send visitors, how you compare with competitors in sampled prompts and whether any of it turns into enquiries. Build it once, annotate the gaps and review it monthly. If you would like our team to set up reporting for you, see our SEO reporting service or request a quote.
References
- Data Studio release notes – Google Cloud
- Connect to Search Console – Data Studio documentation
- What's new in Google Analytics – Google Analytics Help
- [GA4] Default channel group – Google Analytics Help
- Generative AI performance report (Search) – Search Console Help
- New controls and insights for website owners – Google blog (June 2026, updated August 2026)
- Bing's improved AI Performance section – Practical Ecommerce (June 2026)
- Create an AI Brand Visibility Report – Semrush Blog (September 2026)
