Why Google Search Console Impressions Dropped in September 2025

TL;DR:
In September 2025, I noticed the same unusual pattern across several websites: Google Search Console impressions dropped sharply while average position improved, despite clicks remaining relatively stable. The change coincided with Google removing the ability to display 100 search results using the &num=100 parameter, which disrupted rank-tracking tools and appears to have removed a large amount of automated impression data from Search Console. Google later disclosed a separate Search Console impression logging issue affecting data between May 2025 and April 2026, so historical impression and average-position data from this period should now be interpreted with caution.

Each month I carry out an audit of site and search performance for clients who I provide performance insights support for.

For each one, in September, I noticed a strange pattern in search performance.

Why I noticed something strange in Search Console

Each month I carry out site and search performance audits for clients I provide performance insights support for.

In September 2025, I noticed the same strange pattern appearing across several of them.

Search impressions were dropping, while average position was improving.

At first, this didn’t make much sense. If the sites were seeing an improvement in rankings, why were impressions falling?

The following shows the pattern across my own website and three client sites:

What made it even stranger was that clicks hadn’t changed dramatically either way.

Google SERPs are constantly changing through the introduction and expansion of shopping results, AI Overviews, video, images and other search features, so initially I wanted to know whether another Google update had changed what users were seeing.

That’s when I discovered that Google had stopped supporting the &num=100 search parameter.

What changed with Google’s &num=100 parameter?

Around 10-15 September 2025, Google stopped supporting the &num=100 URL parameter in its search results.

The parameter had been commonly used by SEO tools and rank trackers to request 100 Google search results on a single page, rather than retrieving the standard number of results.

That mattered because automated tools could effectively load search results containing URLs ranking much further down the SERPs. If one of your pages appeared in those results, that activity could contribute to the impression data reported in Google Search Console.

Once &num=100 stopped working, SEO tools also began reporting disruption to their rank-tracking data. At roughly the same time, many site owners noticed significant changes to Search Console impressions and average position. Contemporary reporting documented both effects, although Google didn’t confirm that removing &num=100 was specifically responsible for the Search Console changes.

Why did Search Console impressions drop?

The most obvious explanation at the time was the change to Google’s &num=100 parameter.

If automated rank-tracking tools were previously loading search results containing 100 listings, URLs ranking much further down the results could potentially generate Search Console impressions even though no human searcher had actually scrolled through that many results.

Removing the ability to request 100 results at once would therefore reduce that automated activity, particularly for queries where a website ranked well beyond the first page.

That explanation also fitted what I was seeing in the data: a substantial reduction in impressions without a corresponding reduction in clicks.

However, there’s an important update to this story.

In 2026, Google disclosed that Search Console had also experienced an impression logging issue affecting data between 13 May 2025 and 27 April 2026. Google said this affected impressions and metrics calculated from impressions, including CTR and average position, while clicks were unaffected.

That means I wouldn’t now attribute the September 2025 change exclusively to &num=100.

What remains useful about the original observation is the diagnostic pattern: if impressions change dramatically while clicks and actual organic traffic remain relatively stable, don’t immediately conclude that your search visibility has collapsed.

Look at what changed in the underlying measurement before assuming something happened to your rankings.

Why did average position improve?

Google Search Console calculates average position using the positions recorded when your site receives impressions.

That means the figure can improve even when your actual rankings haven’t.

As a simplified example, imagine your site recorded impressions at positions:

3, 8, 12, 75, 93 and 97

The average position across those impressions is 48.

If the impressions at positions 75, 93 and 97 disappear from the dataset, you’re left with:

3, 8 and 12

The average position is now approximately 7.7, despite none of those three rankings having improved.

This is why the combination I was seeing across several sites was so interesting. Impressions had fallen, average position had improved, but clicks hadn’t changed anywhere near as dramatically.

At the time, the &num=100 change provided a logical explanation for that pattern because removing large numbers of low-ranking impressions would naturally improve the average position being reported.

We now also know that Google’s subsequent Search Console impression logging issue affected average position because the metric is calculated using impression data.

The important takeaway is that an improvement in average position doesn’t always mean your rankings have improved. You need to look at the underlying queries, clicks, impressions and actual organic traffic before drawing that conclusion.

How can you tell whether you’ve actually lost search visibility?

A drop in Search Console impressions doesn’t necessarily mean fewer people are finding your website.

Before treating it as an SEO problem, I look at several signals together.

Check clicks as well as impressions

If impressions have fallen sharply but clicks remain relatively stable, that’s very different from seeing both metrics decline together.

A genuine loss of important rankings would normally be more concerning if it also resulted in fewer organic clicks.

Look at your important queries individually

Site-wide averages can hide what’s actually happening. Check the queries and landing pages that matter to the business and compare their positions, impressions and clicks before and after the change.

If your important rankings remain relatively stable, a large site-wide impression drop may be coming from lower-value or low-position queries.

Check organic traffic in GA4

Compare Search Console with your Google Analytics (GA4) data.

If Search Console impressions have collapsed but organic sessions, enquiries, sales and other important outcomes remain stable, that’s a good reason to investigate the reporting change before concluding you’ve suffered an SEO visibility loss.

Check whether the change affected everything at the same time

Sudden changes across multiple unrelated pages, query groups or even different websites can point towards a wider reporting or SERP change rather than something specific to one website.

That was one of the biggest clues in this case. I was seeing essentially the same unusual pattern across several client websites at approximately the same time.

Don’t rely on average position alone

As the example above demonstrates, average position can improve simply because lower-position impressions disappear from the calculation.

Look at the underlying query and page data before concluding that an improving average position means your rankings have actually increased.

The question I ultimately want to answer is not simply “Did Search Console impressions drop?” but “Did this change affect the search traffic and outcomes that actually matter?”

An important update about Search Console impression data

There is an additional piece of context that we didn’t have when I originally investigated this change in September 2025.

In 2026, Google disclosed that a Search Console logging issue had affected impression data between 13 May 2025 and 27 April 2026.

According to Google, the issue affected the way impressions were logged. Because metrics including average position and click-through rate are calculated using impressions, those figures were also affected. Click data itself wasn’t affected.

That period includes the September 2025 change documented in this article.

It doesn’t mean the &num=100 change was irrelevant. The removal of the parameter happened at the same time, disrupted rank-tracking tools and provided a plausible explanation for some of the changes being observed.

But with the benefit of what Google subsequently disclosed, I wouldn’t now say that &num=100 was definitively responsible for the entire Search Console pattern I saw.

If you’re analysing historical Search Console performance across this period, it’s therefore worth being cautious about drawing conclusions from changes in impressions, CTR or average position alone.

The clicks and actual organic traffic generated by the site provide important additional context.

In short

What originally caught my attention was an unusual pattern appearing across several unrelated websites at the same time: Search Console impressions dropped, average position improved, but clicks remained comparatively stable.

At the time, Google’s removal of the &num=100 parameter provided a logical explanation. We now know there were additional issues affecting Search Console impression reporting during the same period.

The broader lesson still stands.

Don’t diagnose SEO performance from a single Search Console metric.

If impressions suddenly drop, look at clicks, individual queries, landing pages, rankings, organic traffic and conversions before concluding that you’ve lost search visibility.

Sometimes the thing that’s changed isn’t your SEO performance. It’s the way the performance is being measured.

---