Back in December 2012 I wrote a post about how Google was making big strides in the right direction in regards to tackling domain crowding. While this was a positive step for sure, and the data clearly proved that the issue had been dampened, it remained a pretty big topic in the SEO world and many thought that the big G hadn’t done enough.
Yesterday though, I started to see some chatter on Twitter that there may have been a new update looking at this specific issue. Jason Mun tweeted that he was seeing big changes in the way Google handled “site:” queries - limiting a normal search to 3 pages of results:
Can you guys do a site: search in Google for any large website. Do you get more than 3 pages of results?
— Jason Mun (@jasonmun) May 21, 2022
So, got confirmation from the US and AU that doing a site: search in Google returns only 3 pages of results. Anyone in the UK can confirm?
— Jason Mun (@jasonmun) May 21, 2022
Justin Briggs also confirmed this and followed up with a good blog post covering what he was seeing.
Fast forward 24hrs and we finally have data from yesterday to start running some analysis on. Also - one of the first things I do in the morning is check the various SERP trackers for signs of change and this is what I was greeted with on SERPmetrics today:
Woah, big change! My first thought was that maybe this was actually the Penguin update that Matt Cutts has been talking about recently, but on further inspection, page 1 flux was barely registering - which would make perfect sense considering my previous analysis showed that page 1 domain clustering had been minimalised already:
The Process
This time, I looked at a smaller keyword set than last time (2000 compared to 5000 before) but the process was exactly the same.
I compared May 21st with May 14th (7 days comparison) and again considered any sequential clusters of 2 or more results from the same domain as a crowd. For each domain crowd I recorded the start position, the finish position and the number of results spanned. I didn’t look specifically at pages and, so, domain crowds could span them.
The Results
Over the test period, occurrences of domain crowding (2 or more sequential results from the same hostname) fell from 19,107 to 15,526. That’s a change of:
The actual number of results within a domain crowd fell from 62,224 down to 49,922 – a similar drop of -19.77%. As before though, most of the change happened outside of page one so the impact the general public see will be minimalised. Here’s what the change looks like across pages 1-10 (page 1 saw -0.09% change so doesn’t even register in this chart):
And here’s some idea of the crowd size change (number of sequential results within each cluster). The top chart shows actual number of occurances for each cluster size and the bottom chart shows % change for each cluser size:
Unlike the last domain crowding update, this one seems to impact clusters of 5+ results a lot more heavily than before.
Things seemed to get better for a couple of weeks after May 22, but It sure seems like this got rolled back in the last couple of days. I see eBay taking up the first 5 or 6 slots for many searches again.
Is this update all about domain repeatation for same query in SERP. I am not getting this clearly.
Hi Shubham - yes you’ve nailed it. Domain crowding/clustering is when you see multiple consecutive results from the same root domain.
Pingback: Penguin 2.0 Reflection: Google Getting Back to Its Academic Ways | OpenView Blog