Google's Search Snippets Go Longer:
What We Found Out Researching 1M Snippets

Michael Konstantinov

Data Science Specialist at Serpstat

Unless you were living under the rock for the last weeks, you already know that Google increased the length of the snippets in the search results. We decided to run a research to find out how long the new snippets are. In this article, I'll tell you:

what results we've got analyzing 1,000,000 snippets;

whether you should change the length of descriptions (spoilers: you should);

how to do it properly.

What's changed?

Let's compare the snippet of the same page shown for the same query before and after this Google update. Let's type some query into the search bar to see how snippets used to look like and how they are displayed now. All Star Wars fans are eagerly waiting for Episode 8 –The Last Jedi to hit the big screen, thus I decided to search for "Star Wars."

Using Serpstat's "Top by Keyword" tool I found how the SERP looked like before this update. The second organic result on September 12 was Wikipedia, let's take a closer look at its snippet. As you can see the snippet's length was 165 characters.

Now let's search for the same query to see what has changed. Without even counting the characters, it's easy to see that the snippet length increased dramatically. Instead of regular 2 lines, there are 4 now. Let's now count the number of characters: 325 characters are shown in the snippet.

How did we run an analysis and what results did we get?

We took 10 000 HV eCommerce keywords from our US database and analyzed their top-100 search results. Thus, we got 1 000 000 snippets to analyze. Here's the histogram we created based on the snippet's length distribution:

Have you noticed how similar this graph resembles two top New York City sights? Brooklyn Bridge and Empire State Building!

As you see, the graph we got is quite interesting as it's bimodal distribution, let's call it "twin peaks distribution graph." Thus there's no sense to count the average snippet length for this kind of distribution. The classic approach to an analysis of bimodal distribution is to represent them as a mixture of two normal distributions.

As you might have already guessed, the left part of the distribution consists of the old and small snippets, with an average length of 140 characters. The vast majority of old snippets count between 80 and 150 characters.

The first peak covers the webpages that:

1) were optimized for 140 characters, which was a recommended length;

2) weren't re-indexed by Google for a long time (ago in a galaxy far, far away).

While the second peak illustrates long and updated snippets, with an average length of 295 characters. The vast majority of new snippets count between 270 and 320 characters.

I also want to share the snippet with the maximum length, which is 345 characters!

Is it time to make your descriptions longer?

Despite the fact that descriptions aren't a ranking factor and making your descriptions longer won't affect your site's ranking, you can still profit from this.The well-written description that includes relevant keywords will attract users and make them click through to your page, thus CTR will grow and it'll positively affect behavior factors of your website.

Moreover, during the latest webmaster hangout (Dec 12) John Mueller — Webmaster Trends Analyst at Google suggested updating your meta descriptions to support the longer search results snippets. Watch this video here:

We're increasing the recommended description length at our
Audit module from 100-140 characters to 300 characters.

What to write in updated description?

If Google considers that your description is not relevant to a user's query, it will form a snippet based on your body text. Google highlights in bold not only the keywords you've typed to the search bar, but also the LSI keywords within this niche. That's why it's highly important to include relevant LSI keywords both to description and body text.

To be sure that you use only relevant keywords within your niche, try our Text Analytics tool. It provides you with a list of recommended keywords to use in Title, H1, and Body based on top-10 search results for all keywords. Moreover, our Text Analytics tool operates in a similar manner as Google thus you use it to easily get the LSI keywords within your niche.

Keyword Clustering. Part 2: Text Analytics

That's it, go updating your descriptions ;)

Rate the article on a five-point scale

The article has already been rated by 2 people on average 4.5 out of 5

Found an error? Select it and press Ctrl + Enter to tell us

Google's Search Snippets Go Longer: What We Found Out Researching 1M Snippets

What's changed?

How did we run an analysis and what results did we get?

Is it time to make your descriptions longer?

What to write in updated description?

Google's Search Snippets Go Longer:
What We Found Out Researching 1M Snippets