It’s common to hear SEOs discuss the “increasing dominance” of big brands in SEO, and how smaller companies just can’t break into the rankings like they used to. Google even put out a “domain diversity” algorithm update a couple of years ago to address the issue, and people like me have shown time and time again how metrics like branded search volume and domain authority — typically signs of a big, well-known company — are key predictors of search performance.
This post, though, is to share some surprising data we’ve surfaced at Moz suggesting that, actually, right now is the best time in years to be an outsider in SEO.
What is domain diversity?
I think there are two appealing ways to define domain diversity, and we’ll look at data for both.
Normally, “domain diversity” is used to refer to a greater number of unique sites appearing in a given SERP. For example, if you have three results from Pinterest and two from eBay on the first page, that is an example of very low domain diversity. I’ll talk about this as “per-SERP” domain diversity below.
The other type of domain diversity I’d like to discuss is the extent to which the biggest sites dominate SEO in general. If the same few sites are ranking top 5 for any query you can think of, that doesn’t necessarily mean low per-SERP domain diversity, but it is still a very homogenous and inaccessible search landscape. I’ll talk about this as “overall” domain diversity, but this is also the domain diversity metric shown on MozCast.
Per-SERP domain diversity
For this measure, we’re using the percentage of page one results which are the first appearance of that subdomain. So for example, if the first page of results has 10 organic listings, of which eight are unique but two are en.wikpedia.org, that’s a score of 90% (i.e. 9/10). Another way of thinking about this same metric is the average ratio of unique subdomains (9) to total subdomains (10). The dataset used throughout is the MozCast corpus.
Here’s how that looks for the last four years, up to August 2021:
To my eye, this is almost incredibly level. For virtually the entirety of 2019, 2020, and 2021, we’ve hovered between 90 and 92%. That’s roughly equivalent to each SERP having one duplicated subdomain. I’ve also included the same statistic if we count sitelinks, which obviously is a little lower as sitelinks always repeat the main site domain, but this is still remarkably consistent.
There’s a five-day dip between October 4 and October 8, 2019, and a longer period of fluctuation starting June 28, 2018, but neither line up particularly closely with any major algorithm update or SERP feature change, and both were ultimately corrected.
Google’s own Site Diversity update on June 6, 2019 barely registers, with a 0.5% impact on the day: