Reading the Performance report

Module 2: Search Console, Bing and IndexNow · Reading time: 13 minutes

Learning objective: After this lesson you can read the Search Console Performance report, explain the four figures correctly, filter and compare the data, and turn the figures into a concrete opportunity for improvement.

Key concepts: Performance report, clicks, impressions, CTR (click-through rate), average position, queries, pages, countries, devices, search type, date range, compare, filter, anonymised queries, fresh (preliminary) data, export (1,000 rows), branded queries

The four figures

The Performance report for search results (in the menu under Performance) shows how your site appears in Google Search. At the top there are four tiles. Click them on or off to show them in the chart.

A few details make the difference between reading the report correctly and getting it wrong.

An impression is not a visitor. An impression counts when your link was on the page of results the searcher was shown. Whether they actually looked at it, nobody knows. Two results from your site for the same query count as one impression in the chart.

Average position is an average. Position 1 is the top result. If your page appears at position 2 for one query and at position 40 for another, the average is 21: a number that fits neither query. So look at the position per query in the table, not just at the total at the top. In the chart the average covers the whole site; in the table it is per row.

Clicks in Search Console are not visits in your statistics. An analytics tool (lesson 8.1) counts differently: it misses visitors who block tracking and also counts visitors from other sources. So don't expect the numbers to match.

The table: queries, pages and more

Below the chart is a table with tabs: Queries, Pages, Countries, Devices, Search appearance and Dates. Click a row in the table and it becomes a filter. That way you can see, for example, which queries one page appears for, or which pages Google shows for one query.

At the top you set:

• Search type: Web (text, and multimodal search such as Google Lens separately), Images, Video or News. Web is selected by default.
• Date: the last three months by default. Search Console keeps sixteen months; if you want to look back further, export regularly.
• Compare: two periods side by side, for example this month against the same month last year. For a seasonal business such as a florist, that is the fairest comparison: a month before Mother's Day is always busier than the month after.
• Filters on query, page, country, device and search appearance, with "containing" or "not containing".

The most recent days are preliminary: they can still change over the next few hours (a dotted line in the chart). Don't draw conclusions from the last two days.

Why the table doesn't add up

If you add up the clicks in the Queries table, you almost always end up lower than the total in the tile. That is not a mistake. Google leaves out queries made by too few people ("not issued by more than a few dozen users over a two-to-three month period") to protect searchers' privacy. These anonymised queries do count towards the total at the top, but they don't appear as a row in the table. If you set a filter on query, they also drop out of the total.

For a small, local site that can be a large share: many people search in their own words, and each variation only occurs a few times.

One more limit: on screen and in the export you see at most 1,000 rows. For most small sites that is enough. If you need more, use the Search Console API.

From figures to action

The report is only useful if you act on it. Four patterns you will almost always come across:

1. Many impressions, average position between roughly 5 and 15, few clicks. Google finds the page relevant, but you are just short of the top or your result doesn't invite a click. First look at the title and meta description (lesson 3.1): do they promise an answer to this query? Then at the page itself (lesson 7.3).
2. Many impressions, very poor position (for example 30 or lower), hardly any clicks. Your page shows up somewhere deep in the results for a search term that only half fits you. Often a search term that is too general ("pioenrozen", peonies, for a local florist). Not every impression is an opportunity.
3. High CTR for your own name. Queries containing your business name (branded queries) almost always score high: people were looking for you. Remove them with a filter "bevat niet: lotte" (not containing: lotte), otherwise you can't see how you are doing with people who don't know you yet.
4. A page that suddenly drops. Compare the period before and after. If only that one page drops, check the Page indexing report and URL inspection (lesson 2.4) to see whether something changed, such as a noindex or a redirect.

What you should not do: set a fixed "good CTR" as a target. How many people click depends on the query, on ads and maps above your result, and on the device. Compare your pages with themselves over time, not with a figure from the internet.

In practice: Bloemenatelier Lotte

Lotte looks at three months and sees, among other things, in the Queries table:

"bloemenatelier lotte" is her name: fine, but no growth. "bloemist amersfoort" (florist amersfoort) is shown often at a position of around 7: this is her biggest opportunity. She clicks that row, chooses the Pages tab and sees that Google shows her contact page for it, not the homepage. She rewrites the title and description of her homepage and makes sure that page clearly says she is a florist in Amersfoort. "boeket bestellen" (order a bouquet) she leaves alone: those searchers want a national online shop.

Bing Webmaster Tools has a similar report (lesson 2.5), with its own name and layout. The concepts of clicks, impressions, CTR and position are the same there.

Summary

• Clicks, impressions, CTR and average position: read them per query or page, not just as a total.
• Average position is the average of your highest result and can be misleading for very different queries.
• The table doesn't add up to the total because of anonymised queries; screen and export show at most 1,000 rows.
• Compare periods fairly (same season), filter out your brand name and ignore the preliminary last days.
• The best opportunities: many impressions at a position just below the top, with few clicks.

Quiz

Exercise

Exercise A2.3: Reading an export. Download the sample file prestaties-zoekopdrachten.csv: a three-month export from Bloemenatelier Lotte. Answer the five questions in the exercise (most impressions, calculating a CTR, the biggest opportunity, the number of branded clicks, and why the total in the chart is higher). The program checks your answers against a fixed key.

Exercise D2.3: Your own figures. Open the Performance report for your own site, choose the last three months and filter out your business name. Note the three queries with the most impressions, with their position and CTR, and choose one you are going to work on. Write down what you are going to change.

Further reading

• Google, Performance report (search results): https://support.google.com/webmasters/answer/7576553?hl=nl
• Google Search Central Blog, A deep dive into Search Console performance data filtering and limits: https://developers.google.com/search/blog/2022/10/performance-data-deep-dive