Study provides little more than anecdotal evidence.
You’re going to see some headlines today about an SEO study commissioned by The Domain Name Association (DNA). Some of these headlines might suggest that new top level domain names get some sort of search engine boost over .com domains. It’s not quite that simple.
Here’s the deal: DNA commissioned a search engine optimization study (pdf) about new top level domain names. It cherry picked a few niches that it felt would have a good mix of websites built on new top level domain names. It further highlighted four domain names built on non-.com domain names that are ranking exceptionally well.
And although the group was careful to not fall into the same trap Rightside did when Google debunked its headline SEO takeaways, its official headline still reads:
“SEO Expert Research Reveals Search Advantages of Relevant Domain Name Extension”
What the research did find was that the so-called “domain authority” of some of the ranked new top level domain names was much lower than .com domain names ranked for the same terms. This would suggest that it’s easier to rank a site on a new TLD than on .com.
What you won’t see in the material is that the sample size was quite small–about 300 “newer” domains (about 2,000 total).
The report admits:
Since the .com websites had by nature more competitors over time, it makes sense that the domains with those TLDs would have to have a higher domain authority to beat out other websites on the same TLD across a sample where only 10 keywords can rank per page of results.
I’m not disputing the work that the SEO experts did to gather the data. I just don’t think people should swallow the headline and pitch as it is.
There was some bias in how the study was set up. And, had the study shown the opposite, you can bet it wouldn’t have seen the light of day.
Now, could new top level domain names with relevant extensions rank well in Google? Could they actually get some sort of boost? I have no doubt that having a domain name with relevant words in it can provide a search engine boost. There’s some debate about whether this includes the keyword in the domain extension. But it’s plausible, especially when people include it in the anchor text they use to link to the site.
I’ve thought about how you could devise a fair study about search engine rankings across top level domain names. It would be rather hard. You’d need to pick a bunch of sets of two domain names (one .com, one non-.com) and get the same backlinks for them. Both would need to have similar keywords in them, so perhaps somekeywords.com and some.keywords. You’d also need the same content on each domain name…but then you’d run into duplicate content issues with Google.
If someone has a suggestion on how to do this, I’m all ears.
Update: Elliot Silver asked Google’s John Mu about the study’s infographic. Here’s what he said:
@DInvesting That looks misleading. New TLDs can rank well, of course: all TLDs can! Also, Google doesn't use DA for ranking.
— John ☆.o(≧▽≦)o.☆ (@JohnMu) April 25, 2017
Of course, the idea behind “domain authority” is that it encapsulates some of the factors in Google’s algorithm.
Andrew, I don’t think at all that Google is giving a boost to the New gTLDs. At the beginning, all TLDs have just the same chance to rank well. I don’t dispute that at all.
But, when it comes down to it, Google really has no direct control over how people link to each other. What anchor text that they use in the links.
Google’s algorithm still favors links.
You can’t dispute the fact that it’s easier for you to get someone to link to you with your preferred anchor text if you have a keyword rich domain name. So, if you have “red widgets” in your domain name, either redwidgets.com or whatever.redwidgets, you’re going to get links with “red widgets” in the anchor text when people link to you.
When it comes to SEO, I prefer to use a domain with the anchor text before or after the dot. With most “good” keyword rich domain names being harder to come by, it’s easier to use a keyword rich TLD domain and get it to rank well.
Andrew Allemann says
I agree. I think this is why having relevant keywords in your domain can matter. It’s very difficult to devise a study that would show if there’s neutrality between keyword1keyword2.com and keyword1.keyword 2 all else being equal.
Eric Lyon says
I agree, it’s impossible to test on a level/identical playing field due to duplicate content, backlink placements, etc. One of the developments will be tossed into the supplemental index never to be seen, rendering test results invalid/bias.
In early launch of new gTLD’s there was alot of conversation talking about all the negatives of using the new G’s, one thing that was thrown out there in the beginning was , they wont rank as well as a legacy tld. I think message the dna wants to convey , as far seo goes , that you are not at a disadvantage with the new strings.
Andrew Allemann says
That is the message, but not the message they pitched.
Brad Mugford says
Right, this is the wording they used –
“.COMs and traditional extensions have to work harder to rank on the first page of Google results… non .COMs need less Domain Authority to rank well for the same keywords.”
They are completely overselling the results of this “study”.
I agree, Phil, the goal is to make sure people realize that the TLD doesn’t matter in search.
Brad Mugford says
Isn’t it amazing these “studies” always seem to end up in favor of the party commissioning the study.
Calling something a “study” should require a certain amount of objectivity and lack of bias.
Google once again called out a new GTLD “study” for misleading or bogus claims.
Chris Boggs says
Hello this is Chris Boggs thank you Andrew for the coverage! I have to admit I was mad/sad when I saw the headline but I see your perspective. PR team is helping The DNA promote the study so we cannot fault them for that. I also had struggled with the study title but as others have pointed out above in the replies, the goal was more to prove that these domains can compete if proper SEO has been implemented.
The full report available to DNA members goes into some additional reasons each of the 4 featured domains has strength. However, the data pulled by Kevin Rowe shows 2k total domains across the results with 300 “non.coms.” those sites had an average DA of about one tenth the traditional sites in the same results. Whether Google wants to argue they don’t use DA or not, the metric is a sound relative data point based on the likely PageRank of the individual pages. The entire industry used this metric, and companies like Moz and Majestic have spent millions in the R&D.
The goal was to show the “new” TLdomains can rank alongside the old ones and they did, but the DA findings are definitely important. Also, there are a number of other nuggets of SEO reasoning in the full report so the headline of your article still hurts.
Again thanks for the coverage!
Andrew Allemann says
Hi Chris, I made sure to point out I don’t doubt the research is solid.
I also agree about DA. Obviously, Google doesn’t “use” DA in its algorithm. The idea is that DA encompasses what Google considers important.
Kind of ironic they were so busy trying to trump up implied benefits that they neglected to mention the primary legitimate benefit: increased likelihood of earning keyword-relevant anchor text. If you did this as a more comprehensive study looking at a lot more ranking factors, I would bet the results would be pretty much identical to our study on keyword domains: http://www.canirank.com/blog/keyword-domains/ For SEO purposes at least, it doesn’t really matter whether the keyword comes to the right or left of the dot.
Chris Boggs says
Matt I don’t really have time to check out your link, but if you were a DNA member you would see much more information in the full member study. I did discuss the value of the anchor text implicitly, and also covered that the keyword could be anywhere in the URL and help for SEO.
People think they can get the truth from salespeople?
Do you believe they would have published a study showing their TLDs don’t work for SEO?
Ozair Akhtar says
I have seen one website which is ranking on Page no.1 of Google though domain TLDs might not be its reason. But this post and other info clarifies that if you choose a niche related TLD then it doesn’t mean that it will take a long for it to rank or even it won’t rank at all. Both .com and other TLDs have a good chance of Ranking.