What 20,000 keywords say about Google’s first page
The number of first-page results on Google has changed from 10 to 8 or 9. In this analysis of 20K keywords, I explain 4 important relationships to better understand the change.
Google’s first page no longer shows 10 results by default on desktop. In 10 search results are no more, I showed that SERP Features like video carousels decrease the number of Blue Links. As a result, there is less playground to compete and the top 5 positions matter even more than before.
My analysis was manual and done at the hand of 44 keywords. Today, I present the results of my analysis of 20,000 keywords! Similarweb was so kind to provide me a ton of data, so we can understand what’s really going on with page one.
A few facts about the keyword set:
No brand keywords
20K ecommerce keywords
All keywords in English
Results from Google US, Desktop
Keep in mind that the landscape on mobile looks very different since Google rolled out continuous scrolling in mobile search. I also noticed that there was a good amount of fluctuation of the number of first-page search results. [x]
When talking about “organic results”, I mean Blue Links. You could argue that SERP Features like PAAs or Carousels are also organic results, but in this context is mean organic snippets. If you’re about the impact of SERP Features on organic traffic, read The impact of image SERP Features on traffic.
To best understand how to think about page one on Google, we have to understand 4 relationships.
Number of results
The first relationship is between keywords and number of results. The number of results Google shows on page one is not always consistent.
The set of 20,000 keywords shows that only 38% first-page search results had 10 Blue Links. Most SERPs actually show 9 organic results (43% of the time) and only 6 keywords in my sample showed 5 Blue Links.
And example of a keyword that shows 10 results is teen clothes. A search for hair dyes, on the other hand, provides only 5 results.
You probably wonder how fewer results impact on organic traffic.
Similarweb tracks zero-click searches, meaning how often a search did not result in a click on an organic result. Often, the reason is because Google enriching the search results with direct answer or visuals that make a click to a website redundant. Ths is the second relationship: first-page results and zero-click searches.
In my analysis, I found a negative correlation of -0.456 between zero-click results and the number of first-page results. In plain terms, the more results Google shows, the more they’re clicked.
A keyword like smiley face nails has a zero-click rate of 97%, meaning the majority of searches don’t end in a click. As you can see (screenshot), Google shows an image carousel at the top, which indicates that the user intent here is inspirational. Users don’t need to click through to Pinterest because they already get inspiration from the carousel at the top.
This makes logical sense at first: more results means more clicks. But at second thought, Google could show direct answer for more results just as much. What’s happening here is not causality but a correlation with user intent.
Query length and first page results
The longer a query is, the easer it is to understand user intent. This dynamic is quintessential for the understanding of SEO and the third relationship in this case study.
In the data I found a strong correlation of 0.921 (that’s might strong) between the query length and number of page 1 results. The longer the query, the more results Google shows.
This again, might be counterintuitive at first because you’d think that if a longer query provides Google with a clearer understanding of user intent, it should be able to give an answer with fewer queries. However, when looking at the results we see one of the biggest differences between SEO today and 10 years ago: Google shows more SERP Features for shorter queries.
The reason is that short queries have ambiguous user intent, meaning they’re searched by users with different intentions. To satisfy them all, Google displays SERP Features like the Popular Products module, map packs, and video carousels, which decreases the number of Blue Links.
As you can already imagine, the length of a query also impacts how many people search for it. The longtail theory says that the finer an interests, the fewer people have it.
Search volume and first page results
The fourth relationship that helps us understand Google better is between the number of first-page results and search volume. It has a negative corelation of -0.671, meaning the shorter the query, the more search volume it has.
This makes sense and is nothing new for readers of the Growth Memo, but to see proven over and over again. The longest query in the data set is “bee hive insulation wraps for the winter for sale best” with 278 monthly searches and the shortest “frames” with 33,827 monthly searches.
More insights from the data
I want to close with thoughts on how these insights can make you a better SEO. The biggest takeaway from the data is that targeting shorter keywords is less predictable, more competitive, and demands a SERP Feature-focused strategy. Longer queries provide space for classic Blue Links and competitive advantages from Rich Snippets.
The common advice amongst SEOs is to best go after the longtail if you want to win in SEO but that’s not universally true. Established brands can go after the shorthead, but they need to create content in multiple modalities: images, videos, articles. Within these formats, shorthead keyword strategies need different varieties to target all possible SERP Features: user-generated and highly produced images, long and short-form videos, reviews/tutorials/buying guides and news articles.
My analysis of 20,000 ecommerce keywords will be split up into two articles. The first one (this one) is a deep dive into the mechanics behind first page results. In the second one - coming out next week - I will revise the relationship between SERP Features and first-page results.