Looker Studio SEO dashboard for clinics: GSC + GA4
A Looker Studio SEO dashboard for clinics: two Search Console data sources, GA4 connector quotas, a locale-based page structure and the charts to leave out.
By Roozbeh Nazari · CEO
The purpose of building an SEO dashboard for a clinic is that the marketing lead can look at the data in their own account on one screen, instead of asking the consultant "how are we doing" every month. Looker Studio is a sufficient and free tool for this; the difficulty is not the tool but deciding which data to pull from which source, and what not to chart. This article describes the structure of the dashboard we build for a multilingual clinic site, the limits of the Search Console and GA4 connectors you need to know about, and the reasoning behind the charts that stay out.
The source principle
All of the dashboard's data sources should be tied to the clinic's own Google account: the Search Console property, the GA4 property and the dashboard itself. The consultant builds the dashboard as an editor inside the clinic's account; they do not build it in their own account and share it. This principle ensures that the dashboard and the data history stay with the clinic when the consulting relationship ends. It is also the first rule of our analytics and data service: the measurement infrastructure belongs to the client.
Search Console: you need two data sources
Looker Studio's Search Console connector reflects Search Console's two different aggregation methods: site impression and URL impression. In the site impression table, impressions and clicks are counted at site level; it is the right source for query-based analysis. In the URL impression table the same metrics are counted at URL level; it is the right source for page-based analysis. Google's documentation says it plainly: a single data source can use only one of these methods. If you want to see both, you create two separate data sources for the same property.
The distinction matters in practice because the totals of the two tables do not match, and that is normal. When more than one of your pages appears on the same results page for the same query, site impression counts one and URL impression counts two. If you do not want to answer the question "why are they different" every month after placing numbers from both sources side by side, write which source each chart comes from in its title.
Search Console data has one more limit: the performance report goes back up to sixteen months. You can make a year-over-year comparison but not one against two years ago. If you want a longer history, you need to export the data to a table at regular intervals; that is a matter for the data retention plan, not the dashboard.
GA4: connector and quota
The GA4 connector is easy to set up; the difficulty comes later. Google's documentation says that Looker Studio reports connecting to GA4 are subject to Google Analytics Data API quotas, and that reports exceeding the quota may show an error. The quota is consumed by the query every person who opens your dashboard makes for every chart. If ten people open a thirty-chart page on the same morning, seeing an error is normal. We take three precautions: keeping the number of charts per page around ten, setting data freshness to twelve hours, and gathering the most-viewed metrics on a single summary page while opening detail pages only when needed. The data freshness setting varies by connector type; extending the default for GA4 noticeably reduces quota consumption.
The metrics we bring from GA4 to the dashboard are limited: organic sessions, organic key events (form submission, WhatsApp click, phone click), key event rate by landing page, and a breakdown by locale. We described how to set up the locale breakdown on a multilingual site in our GA4 multilingual reporting article; the dashboard sits on top of that structure.
The dashboard's page structure
Page one, summary: four numbers and two line charts. The numbers: organic clicks (Search Console, site impression), organic sessions (GA4), organic key events (GA4), organic enquiries marked qualified in the CRM (a table entered manually or connected from the CRM). The lines: organic clicks and organic key events over the last sixteen months. This is the one page the marketing lead looks at every week.
Page two, queries: from the site impression source, clicks, impressions and average position filtered to the target query set. The target query set is a list agreed with the consultant and defined as a filter in the dashboard; the aim is to show the queries being tracked, not to list every query. A country filter is mandatory here: combining Arabic queries from the Gulf countries with Turkish queries from Turkey in one table misleads.
Page three, pages: from the URL impression source, clicks and impressions grouped by locale folder, with the same URL's key event count from GA4 alongside. Showing these two sources in one table requires data blending; blending joins fields from different sources on a common key, and the key here is the page URL. Make sure URL formats are the same in both sources; Search Console gives the full URL while GA4 gives the path, and you need to bring them into the same form before blending.
Page four, technical health: Search Console's indexing data does not come through the connector; on this page we place a monthly table entered by hand: submitted and indexed URL counts per locale, and the share of URLs passing the good threshold in Core Web Vitals. Manual entry is not ideal, but it beats leaving the page empty while waiting for a connector that does not exist.
What not to chart
Some charts are dashboard decoration and produce no decisions. The total impressions line: an impression is not an outcome unless it turns into a click; its growth looks good but says nothing on its own. The site-wide average of average position: the position average of hundreds of queries gets worse when you start appearing for a new query; that is expansion, not decline. Device and browser distribution: interesting, but produces no decision. Real-time user count: has no place on an SEO dashboard. Removing these charts keeps the dashboard readable and the quota sustainable. The same principle applies to the monthly report you ask your consultant for; in our article on report content we listed separately which metrics are decoration.
Maintenance
A dashboard is not built once and left alone. The quarterly maintenance list: is the target query filter current, have new locale or treatment pages been added to the URL filter, have key event definitions changed in GA4, are the data sources' credentials still valid? The last item is the one most often forgotten: when the account of the person who set up the data source is closed, the dashboard goes blank. That is why we set up data sources with the clinic's shared admin account rather than a personal one.
Our own practice
With clinic clients we build the dashboard by the end of the first month and use it as the input to the monthly meeting. You can see our public sample dashboards on our reports page. The dashboard's value lies not in the beauty of the data but in looking at the same four numbers from the same place every month; that consistency is what answers the question "what changed" at the end of the year.