Google Search Console Dashboard: Five Panels Worth Building, and the Arithmetic Behind Them

A useful Search Console dashboard is five panels and a rule about ratios. Most of the ones that mislead do so because they average a CTR column.

A good Google Search Console dashboard has five panels: a clicks and impressions trend, CTR and average position, query movers, page movers, and a note on which days are still preliminary. Build it from summed totals, not from averaged columns, and compare weeks or months against the previous period and last year. You can assemble it by hand from the Performance report or Looker Studio, or you can connect Search Console to Quiriz and ask for each panel in plain English. Below is the layout, the arithmetic, and the limits of the data.

The five panels

1. Trend. Clicks and impressions by week for the last 13 months. Weekly smooths out the weekday effect; the 13 months give you a year-over-year comparison for every point.
2. Efficiency. CTR and average position, shown by month. These are ratios, so they get their own panel and their own rule (next section).
3. Query movers. The queries whose clicks changed most against the previous period, up and down. Show the change, the current clicks and the impressions.
4. Page movers. The same table for pages. This is the panel that tells you which URL to open first.
5. Freshness note. A line saying the newest two or three days are still being finalized, and that days are in Pacific Time. It prevents the most common false alarm: a dip in the last few days that disappears by itself.

Notice what is not on the list: a top-ten queries table sorted by total clicks. It looks like a dashboard and says nothing, because the same ten rows lead every month. The movers are where the information is.

Three ways to get the data onto a screen

The Performance report itself. Search Console's own report is a fine dashboard for one person: set the date range, add a comparison, switch to weekly granularity, and save the filtered view in your browser. It does not combine panels, cannot be shared as a single page, and the table shows at most 1,000 rows.

Looker Studio. Google's free reporting tool has a Search Console connector, so you can lay out panels, share a link and have it update. You own the chart layout and every calculated field, which is where the next section matters.

A spreadsheet. Export, paste into a sheet, build pivots. It works, and every month you rebuild it. The SEO report template walks through the export step by step and the limits you meet there.

The one rule: never average a ratio

CTR is clicks per impression. Average position is where you ranked, weighted by how often you were shown there. Neither can be summed, and neither should be averaged across rows.

For any period, CTR is the sum of clicks divided by the sum of impressions. Average position is the sum of position times impressions, divided by the sum of impressions. The numbers below are an example, not a benchmark:

The simple average says 7.0% and 4.5. The right answer is 4.1% and 5.9. A dashboard with the first row of arithmetic on it will report an improvement that is not happening whenever small, high-performing pages grow faster than large ones.

Questions the dashboard should be able to answer

• Did total clicks fall because impressions fell, or because CTR fell? (Different causes, different fixes.)
• Which five pages lost the most clicks against last month, and did their position move?
• Which queries have plenty of impressions and a low CTR?
• Is the newest month still moving, or is it final enough to report?

A dashboard that answers these in under a minute is working. A dashboard that needs a spreadsheet beside it to answer them is a picture.

How Quiriz builds it: by asking

Quiriz is an AI business analyst. Its Google Search Console connector is read-only, refreshes daily, and is included in paid plans. You sign in with a Google account that has access to the site, pick one site, and Quiriz reads three datasets over the last 16 months:

• [Search Console] Daily performance: whole-site clicks, impressions, CTR and position per day.
• [Search Console] Queries by month: the top 5,000 queries per month by clicks.
• [Search Console] Pages by month: the top 5,000 pages per month by clicks.

The metrics are defined once: Search Console clicks, impressions, CTR, and average position. CTR and average position are computed from summed clicks, impressions and position-weighted impressions, so the arithmetic in the table above is built in rather than something each chart author has to remember. Each panel is a question:

• How did our Google search clicks and impressions change each month this year? (the trend panel)
• Which search queries brought the most clicks last month? and Which queries have many impressions but a low click-through rate? (query panels)
• Which pages got the most clicks from Google search last month? (page panel)

You can ask for the movers directly, for example the pages whose clicks fell the most against the month before. Because the metric definitions live in Company Context, an admin can also define something like branded clicks once, as queries containing your brand terms, and everyone who asks gets the same number. The answer shows its work, so you can see which dataset and filter produced a figure. The data refreshes daily, so asking the same question tomorrow reads the newer numbers. For agency teams running this across clients, see Quiriz for agencies.

Limits worth knowing

• Days are Pacific Time and the newest two or three are preliminary. Google adds them once final, so the last few days will move.
• Query rows leave out anonymized queries. Google omits rare queries for privacy, so the queries dataset adds up to less than the daily totals.
• Page rows do not add up to the daily totals either. Search Console counts each page separately, since one search result can show several pages from your site, so page impressions can be higher and clicks lower than the daily figure.
• Only the top 5,000 queries and pages per month are kept. A small site will never hit that; a large one will lose its long tail.
• One site per connection, and the dataset has no country, device or query-by-page breakdown. The three tables above are what you can ask about.

If you also need conversions, the companion post on pages that rank but do not convert joins this data to GA4, and finding declining pages goes deep on the movers panels.

Frequently asked questions