Chapter 29: Reading Google Search Console Correctly

Last updated: November 1, 2026.

One of the biggest mistakes in SEO is looking at data and jumping straight to a conclusion.

Google Search Console gives us very valuable information, but if we don’t interpret that data correctly, the same tool can lead us to the wrong conclusion.

In this chapter, we’ll look in particular at:

• Clicks
• Impressions
• Click-through rate (CTR)
• Average position
• Query-level analysis
• Page-level analysis
• Country, device and date breakdowns
• Branded / non-branded queries
• Regex filters
• URL Inspection
• Indexing reports
• Core Web Vitals
• Situations where Search Console data gets misread

I think the most important sentence about Search Console is this:

Seeing the data is one thing. Interpreting it correctly is another.

1. What does Search Console tell us?

Google Search Console is one of the most fundamental tools for seeing how your site performs in Google Search.

In the Performance report, we see four main metrics:

• Clicks
• Impressions
• CTR
• Average position

We can also break the data down by:

• Queries
• Pages
• Countries
• Devices
• Search appearance
• Dates

The biggest advantage here is that we see data about Google traffic directly through Google’s own system.

2. What is a click?

If a user clicks through to your site from a Google search result, Search Console counts it as a click.

For example, for the query “seo agency” your site may have appeared 10,000 times and been clicked 500 times. In that case, you got 500 clicks.

But are 500 clicks good or bad? The answer depends on these questions:

• How many impressions were there?
• Which query did they come from?
• Which page did they land on?
• Which country did they come from?
• Which device did they come from?
• What was the number in the previous period?

The click count alone doesn’t give us enough information.

3. What is an impression?

An impression can occur when your page is shown to a user in Google’s results.

I think impressions are a very important early signal. Sometimes impressions start rising before traffic does.

For example, say you published a new piece of content:

Clicks may not be very high yet, but Google may have started showing the page for more queries and to more users. So if we only look at clicks, we can miss the early stage of growth.

4. What is click-through rate (CTR)?

CTR, or click-through rate, shows what share of impressions turned into clicks.

Formula: Clicks / Impressions x 100

For example, if there are 1,000 impressions and 100 clicks, CTR is 10%.

But again, there’s an important point:

A low CTR doesn’t always mean a bad title.

5. Why shouldn’t CTR be interpreted on its own?

Let’s consider two queries.

Query A:

• Average position: 1.8
• CTR: 35%

Query B:

• Average position: 9.4
• CTR: 2%

Seeing the low CTR on query B and saying “The title is bad” may not be right. The user is already seeing you near the bottom of the page. The problem may not be the title but the ranking.

Also, CTR at the same position can change if the SERP has elements like these:

• Ads
• AI Overviews
• Maps
• Videos
• Shopping results
• Featured snippet

That’s why CTR always needs to be read together with position and the structure of the SERP.

6. Average position is one of the most misread metrics

I think this is the most misread metric in Search Console.

Many people think, “Average position is 7.2. So we’re in 7th place.” That’s not actually the case.

An average position of 7.2 can mean you appear in 5th place for some searches and 10th for others. On top of that, you may have different rankings in different countries. For example, you could be:

• 11th in Turkey
• 3rd in Germany
• 7th in the US

The overall average shows us a single number, but if your target market is Turkey, the data you really need to look at is Turkey’s.

Search Console’s country filter can split the data by the country the search was made from. So, especially on international sites, it’s useful to always check average position for the country we’re targeting. Google Help

In short:

Average position is not a fixed rank. It’s a trend and an average.

7. Averages sometimes hide the truth

Let’s say a page gets traffic from two queries:

• First query: position 1
• Second query: position 19

The average may come out around 10. But in reality the page isn’t in 10th place for any query. One is very strong, the other very weak.

That’s why making decisions based on the site-wide average position is usually not enough.

You need to break it down by query and by page.

8. Why can moving from position 11 to position 10 matter?

In SEO, a position change that looks small can sometimes have a very big result. Moving from the second page to the first is one of them.

In various CTR studies, position 11 and beyond usually drops to click-through rates of around 1% or lower. For example, in Advanced Web Ranking’s 2026 data, the average CTR for position 11 on US desktop is about 0.67%. This rate can vary by search type, device, country and SERP structure. Advanced Web Ranking

So moving from position 11 to position 10 isn’t just “we went up one spot.” Because you’ve moved onto the first page, you can see a much bigger change in impressions and clicks.

That’s exactly why I like to track queries in the 10 to 15 range.

9. Why is query-level analysis important?

In the Queries section, we can see which searches users saw us for.

I especially like to look at these kinds of queries:

• High impressions, low clicks
• Average position between 5 and 15
• Queries we didn’t expect
• Brand queries
• Queries with commercial intent
• Queries we’ve just started appearing for

For example, one of your pages might rank 15th for “coffee machine.” But the same page might rank 6th for “automatic coffee machine for home.” If we look at site-wide data, we may not notice this.

10. Why are queries in positions 5 to 15 interesting?

I see this zone in particular as an opportunity area, because Google has already associated the page with the query.

For example, the right improvements for queries in positions like these can produce serious results:

• Position 6
• Position 8
• Position 11
• Position 14

These improvements might include:

• Updating the content
• Adding missing headings
• Adding more internal links
• Improving the title
• Meeting the search intent better

Sometimes strengthening a page that already ranks is more efficient than producing new content.

11. What can high impressions and low CTR tell us?

For example, there are 100,000 impressions and 1,000 clicks. CTR is 1%.

There can be many reasons for this:

• Average position is low
• The title isn’t strong enough
• The description is weak
• It doesn’t fully match the search intent
• There are a lot of ads
• The AI Overview answer keeps the user on the results page
• Brand awareness is low
• You’re appearing for the wrong queries

So when we see a low CTR, we don’t change the title right away. First we ask this question:

Why is it low?

12. Page-level analysis

I consider the Pages section as important as the Queries section.

Here we can look at:

• Pages with the most clicks
• Pages with the most impressions
• Pages whose traffic dropped
• Pages whose traffic rose
• Pages that have just started appearing

Date comparison is especially useful. For example, you can compare “Last 28 days” with the “Previous 28 days.”

If a page’s clicks dropped while impressions and position stayed the same, you can investigate CTR or SERP behavior. But if clicks, impressions and position all dropped, it may be a ranking problem.

13. Looking only at total organic traffic isn’t enough

Let’s say site traffic dropped 20%. You shouldn’t immediately say, “SEO is going badly.”

First you need to look at this:

Where is the drop coming from?

Maybe 80% of the total drop comes from a single page. Maybe demand for a high-volume query declined. Maybe a single country dropped.

That’s why, as much as possible, you need to drill down from site > folder > page > query.

14. Don’t forget seasonality when comparing dates

For example, on a site that sells air conditioners, you compare August with September. Traffic has dropped. That’s not necessarily an SEO problem. Search demand may have dropped.

In the same way, periods like these can affect query volume:

• Holidays
• Ramadan
• Black Friday
• The school term
• Summer season
• Winter season

So sometimes a “This year / Same period last year” comparison makes more sense than “Last 28 days / Previous 28 days.”

15. Why does the country breakdown matter?

We talked about its importance in international SEO.

Total traffic may have increased. But the US may have dropped 30% while India rose 200%. If your target market is the US, you have a business problem even though total traffic looks like it went up.

So the question:

Where is the traffic coming from?

is just as important as total traffic. Average position also needs to be read by country.

16. Device breakdown

Desktop and mobile performance can differ. For example:

• Desktop CTR: 8%
• Mobile CTR: 3%

Why?

• The mobile SERP may be different.
• Ad space may be heavier.
• The title may get cut off earlier.
• Local results may take up more space.
• The ranking may be different.

That’s why I always examine the device breakdown, especially on projects with a lot of mobile traffic.

17. What is Search appearance for?

Search appearance helps us understand which types of search results your site performs in.

Depending on the project, you might see appearances like these:

• Product results
• Rich results
• Video results
• Merchant listings

This area is useful when reviewing the results of schema work.

18. You aren’t fully using Search Console until you learn filters

I think filters are one of Search Console’s most valuable features.

For example, we can create filters like these:

• Query: contains seo
• Page: contains /blog/
• Country: Germany
• Device: Mobile

We can also combine them. For example, we can do much more specific analysis with “Germany + Mobile + /de/.” It becomes far more useful than looking at the total data.

19. Search Console now has a branded and non-branded query filter

We used to rely on regex to separate brand and non-brand queries.

Now Google Search Console offers a Branded and Non-branded filter directly for this.

With this filter, Google can split performance by whether the query contains your brand name. The brand classification is AI-assisted, and Google also states that it can misclassify from time to time. The filter may also not be available for low-impression sites or some child properties. Google Help

This is a very useful development. Branded traffic usually helps us understand users who already know you, while non-branded traffic helps us better understand new user acquisition. Google Help

For example, organic clicks rose 30%. But if the entire increase came from brand searches, our interpretation of SEO growth would be different. If non-branded queries also grew, that’s another story.

20. So do we no longer need regex?

We still do.

The branded / non-branded filter made separating brand queries easier, but regex can be used for very different kinds of analysis.

For example, to find question queries:

^(who|what|how|why|which|how many|where|when)\b

This way we can separate informational question queries.

For queries with price intent:

(price|cost|fee|how much)

For queries close to purchase intent:

(buy|for sale|price|deal|discount|order)

For comparison queries:

(best|comparison|vs|alternative|which)

For local queries:

(istanbul|ankara|izmir|konya|near me)

These are just examples. Every industry may have a different regex set.

On large sites, using regex correctly lets us create segments in a few seconds instead of reviewing hundreds of queries one by one.

21. The branded / non-branded filter and regex can be used together

I think this is now the most sensible method.

For example, we can first select the “Non-branded” filter. Then, with regex, we can isolate commercial queries like price|cost|fee. That gives us a very valuable segment: non-branded + price-intent queries.

Or with “Non-branded + question queries” we can analyze information-seeking users who don’t know the brand yet.

This lets us use Search Console much more powerfully.

22. Why isn’t Search Console data the same as Analytics?

Because the two tools measure different things.

Search Console measures impressions and clicks in Google Search. Google Analytics measures user behavior after they arrive on the site.

So a Search Console click and an Analytics session don’t have to match one to one. We’ll talk about this in more detail in Chapter 30.

23. What is URL Inspection for?

URL Inspection is one of Search Console’s most useful tools. We enter a URL and can see what Google knows about that page.

For example:

• Does Google know the URL?
• Is it indexed?
• When was it last crawled?
• Which canonical was selected?
• Is there a crawl block?
• Is structured data detected?

It’s very useful for understanding the technical status of a single page.

24. The indexed URL and the live test aren’t the same thing

There’s an important distinction here.

The initial information in the URL Inspection tool shows the latest version Google knows about. But the Test Live URL option checks the page’s current state.

For example, yesterday there was a noindex. Today you removed it. The normal report may show the old state. The live test may show the new state.

25. What does “URL is on Google” mean?

This message means the page is in Google’s index and eligible to be shown in search results. But:

It doesn’t mean it will be shown for every query or at all times.

Getting indexed is one thing. Ranking is another. I think beginners in particular need to know this distinction.

26. When should Request Indexing be used?

Request Indexing can be used for a new or significantly updated page.

But you shouldn’t try to manage a site with thousands of pages through this button. For a large number of URLs, a sitemap is the better method.

Also, sending an indexing request doesn’t mean Google will definitely index it.

27. The Page indexing report

It’s very useful on large sites.

We can see in a more aggregated way which pages Google has indexed, which it hasn’t, which it has crawled and which it has discovered.

But again:

Not every non-indexed URL is an error.

28. Don’t panic when you see “Not indexed”

For example, these shouldn’t be in the index anyway:

• A redirected URL
• A canonical alternate
• An intentional noindex
• A 404 page

Seeing “100,000 pages not indexed” in Search Console isn’t a problem on its own.

As we discussed with the Filmler.com example in Chapter 25, sometimes it can be better to deliberately get fewer URLs indexed.

29. Crawled – currently not indexed

Google found the URL. It crawled it. But it didn’t index it.

Saying “Google can’t find the page” here is wrong. Google found it. The problem may be somewhere else.

For example, you can look at:

• Content quality
• Similar pages
• Canonical
• How unique the page is
• Overall value

But there’s no single cause.

30. Discovered – currently not indexed

Here Google knows about the URL but may not have crawled it yet.

On large sites in particular, issues like these can come into play:

• Crawl budget
• Server performance
• Too many low-value URLs
• Generating URLs too quickly

The crawling logic we discussed in Chapter 23 shows up again here.

31. Checking the canonical

This is one of the areas I look at most closely in URL Inspection.

There can be two values:

• User-declared canonical
• Google-selected canonical

We may be declaring one URL as canonical while Google has selected another. If they’re different, you need to investigate.

This can sometimes be a signal of a very important technical problem.

32. The Core Web Vitals report

We discussed this in detail in Chapter 24.

Search Console has a Core Web Vitals report based on real user data. Here pages can be split into these groups:

• Good
• Needs improvement
• Poor

The main metrics:

• LCP
• INP
• CLS

The data here is different from a single Lighthouse test. It’s based on real user experience.

33. Why doesn’t the chart total sometimes match the table total?

This can be confusing too.

You see 100,000 clicks on the chart. When you add up the queries in the table, you get 92,000.

One reason is that some queries aren’t shown in the table for privacy reasons. Search Console can also group data differently across different breakdowns.

So you shouldn’t read a Search Console export like an accounting ledger.

34. Position when the same site has multiple results

Let’s say three results from your site appear for a query:

• Position 1
• Position 2
• Position 3

In some views, Search Console may use the position of the site’s topmost result. When you analyze by page, though, the URLs may be evaluated separately.

This is another point to be careful about when interpreting average position.

35. Be careful with the most recent day’s data

Search Console’s most recent data may sometimes not be complete yet.

So before panicking with “Traffic dropped 40% today!”, you need to check whether the data is complete. This matters especially in daily reporting.

36. How do I analyze a traffic drop?

I usually go from the top down.

First: When did it drop? Then: Where did it drop?

After that:

• Which pages?
• Which queries?
• Which countries?
• Which device?
• Did impressions drop?
• Did position drop?
• Did CTR drop?

For example:

37. How do we track the impact of SEO work?

For example, we updated a piece of content. Afterward we can track:

• New queries
• Increase in impressions
• Change in average position
• Increase in clicks
• Change in CTR

But I don’t look at a single day. It’s healthier to note the date of the change and compare after enough time has passed.

In SEO, you need to avoid confusing correlation with causation.

38. My favorite opportunity analysis in Search Console

I especially like to look at this combination:

• High impressions
• Average position 5 to 15
• Low or medium CTR

Google has already associated us with the query. Then we look at the relevant page:

• Is the search intent right?
• Can the title be improved?
• Are there missing headings in the content?
• Can internal links be increased?
• Is the content outdated?
• Why is the competitor above us?

Sometimes the fastest win comes from here, without producing any new content.

39. Does Search Console show us the whole picture?

No.

Search Console is very powerful, but it only shows a certain part of the Google Search side. We can’t fully see these things here:

• What did the user do on the site?
• Did they buy?
• Did they fill out a form?
• How much revenue did they generate?
• Which pages did they browse?
• Did they come back?

For that, we need Analytics, CRM and sales data. We’ll move on to that side in Chapter 30 and Chapter 31.

Wrapping up

Google Search Console isn’t a tool that gives you an SEO score. It gives us data about how Google sees our site and how users interact with us in search results.

But if we take the data out of context:

• We think average position is a fixed rank.
• We think a low CTR always means a bad title.
• We see every non-indexed URL as an error.
• We think everything is fine when total traffic goes up.
• We look at the global average without isolating the target country.
• We confuse brand traffic with non-branded SEO growth.

I think the most important habit when using Search Console is this:

Don’t make decisions by looking at a single metric.

You need to read clicks together with impressions, CTR with position, pages with queries, countries with the target market, and today with past periods.

Because in SEO, what matters more than the numbers themselves is understanding why the numbers change.

Previous: Chapter 28: How to Earn Strong Backlinks · Contents · Next: Chapter 30: Using SEO Tools Efficiently

This chapter was first published in Turkish on X (October 5, 2026).