Drop in Impressions Signals Major Change in SEO

Author Philip Gallagher
SEO
Performance

SEO Lives: Death of the Top 100

It’s no secret that it’s been an awkward two or so years in the world of SEO. AI Overviews has seen website clicks speed downhill, dragging traffic and lead generation along with it. Even with the exciting changes in AI utilisation and ‘Search Everywhere SEO’, it can sometimes feel like we’ve been somewhat starved for good news.

Unfortunately, it looks like impressions are now following suit. Around the 10th of September, webmasters began reporting struggles trying to view Google’s Top 100 results. On investigation, independent SEO consultant Brodie Clark found this had resulted in “impressions declining dramatically” across the SEO community.

What caused SEO impressions to drop so rapidly?

An ‘impression’ is, essentially, a confirmation that your website appeared on a search result page. It’s a signal that you were in the running to win organic website traffic, with your Click-Through Rate (CTR) indicating your success rate. 

While impressions are hardly the most celebration-worthy measurement of an SEO manager’s achievements, it’s nice knowing that your work is at least getting within scrolling distance of users or potential customers. It also helped lessen the sting of the loss in clicks, with ‘The Great Decoupling’ having seen clicks vanish and impressions expand since the introduction of AI Overviews.

Brodie Clark found that the secret lay with Google’s ‘&num=100 parameter’. This handy feature forces Google to show you 100 results from a search, rather than using pagination to continue generating results as you scroll. 

Scrolling through 100 results, rather than simply reading a top-of-page summary, is pretty much the antithesis to AI Overviews. Needless to say, this old parameter doesn’t quite align with Google’s modern approach to search. This could be why the elite search engine finally disabled the parameter, a fact they confirmed after Clark published his investigation.

Speaking to Search Engine Land, a Google spokesperson stated, “The use of this URL parameter is not something that we formally support.”

This decision has already sent ripples through the SEO community. The ‘Top 100’ has long been a staple of tools like SEMRush, AHRefs, and even Google Search Console. It provides a straightforward summary of how a keyword is performing, making results more accessible for people in the marketing industry and clients who care about their site rankings. 

You may be asking what losing this parameter has to do with website impressions. The answer is that these SEO tools were deploying bots that utilised the &num=100 parameter for searches. This would create 100 ranks, which the tools could then display for marketers to skim through. 

What we didn’t previously appreciate, however, is that these bot searches also generate impressions by doing so. The result was greatly inflated impressions data; it was, in essence, fake data from fake searches. By removing the parameter to generate 100 results, Google has simply taken away the ability for bots to generate impressions for 100 websites via a single page load.

Experts like Clark have been shocked by the “sheer number of impressions” that were supposedly being generated by bots. In one example, he found a website losing 200K+ impressions daily. However, the decline in impressions does not signify a loss of user interest; in fact, without the bots, user interest is what remains.

How can businesses adapt to losing the top 100 SEO rankings?

While this development was certainly alarming, it should be fairly simple for companies to adapt.. Businesses must be sure to:

  • Avoid panicking – The impressions you lost most likely came from bot traffic. You can now take both impressions and click-through rate (CTR) as more reliable metrics for your website’s search performance.
  • Be aware of SEO costs – SEO platforms may adjust their costs to reflect the increased expenditure of generating data on rankings. Providers like SEMrush and AHrefs have been quick to comment on developments, so expect to hear more soon.
  • Stick to the basics – This change reaffirms the importance of committing to achieving top ranks. In modern SEO, this translates to providing unique, efficient, and engaging content while exploring longtail topics and building a brand’s authority. While some ranks will be harder to track, the fundamentals of SEO remain solid.
  • Explore AI and SERP ranks – As important as the fundamentals of SEO remain, the name of the game is now ‘search everywhere optimisation’. That means staying aware of your AI Overview ranks, how you appear on LLM platforms, and how you perform for alternative rankings like People Also Ask, Videos, Images, and so on. Keep in mind that alternative results often appear above rank 1! By widening your net, you’ll ensure you are not only optimising your ranks but also helping your content work for you in other ways.

Is losing the top 100 bad for SEO marketing?

So, what does this mean for SEO managers? Should clients be worried? More importantly, what needs to be done?

On the one hand, removing this false perspective has made impressions more valuable as a success metric. Their decline has also seen click-through rates jump, as the ratio of clicks to impressions is suddenly improving dramatically for webmasters worldwide.

On the other hand, this jarring loss of such a core SEO element will challenge platform providers to secure their place in the market. Google’s default is now a measly top 10, and search tools will have to work much harder to generate data.

Summarising how this could impact management, SEO Insight tweeted out:

“Google has killed the n=100 SERP parameter. Instead of 1 request for 100 SERP results, it now takes 10 requests (10x the cost).”

To put it another way, losing the parameter means SEO tools will have to work harder to get the results they need. With this will come greater costs, which are likely to percolate down to platform users, resulting in more expensive membership packages for SEO tools. Certain providers may also offer limited alternatives, showing data for the top 10 to 20 rankings rather than the full 100.

Stepping back from this whole ordeal, it seems as though the full practical impact of losing the Top 100 metric is yet to be seen. SEO strategies will not need to change to any major degree, but the software we rely on is still reshaping itself. Semrush reassured users, “Top 10 & Top 20 results remain fully intact”, while AHRefs warned that “Rankings beyond the top 10 will show inconsistencies.”

However, this does not necessarily mean a huge changeup in best practices for SEO strategy. The Top 100 has always been more convenient than realistic. In a recent study, Backlinko found that position 10 results have an average CTR that’s around one-tenth of top ranks. Worse still, only 0.63% of Google users click a result from a second page. While it’s nice moving from 79 to 42, or 85 to 29, it’s unlikely to translate into conversions in practice.

Granted, this doesn’t mean that only the top-ranking websites enjoy visibility. Only around 54% of AI Overview result citations match immediate organic search results. You don’t have to have a top-ranking URL to feature in AI Overview searches or LLMs, but this hardly makes up for click activity that remains gated in the Top 10.

Actually getting into the Top 10 will be just as much of a challenge as ever, but without the ‘consolation ranks’ that show early progress. The viability of focusing on intent-driven longtails over high-difficulty short tail keywords also remains viable, though the cost of accessing data for gap analysis may yet increase.

What we have been left with is more brutally honest search data. With more accurate performance metrics, we should have a far clearer impression of how optimisation impacts traffic. 

In the end, Google can change the rules whenever it wishes. Marketers will adapt like always, and we’ll soon see the impressions drop as just another stepping stone in Google’s transformation journey.

Philip Gallagher

SEO Manager