This article speaks in terms of whether you should scrap a site and just 301 backlinks to another if it is performing poorly. But it speaks to what IS performing poorly.
There is also another article more recently written all about SEO Metrics that might serve as a good read after you consider this article’s points.
Why? Because 1. Youre asking someone else…each of these numbers are conjured up by a 3rd party. And 2. Its subjective. So lets do an average?
How many sites are on the internet? 1.5 billion. I cant look up all their DAs. So lets form a couple groups that we don’t pick who is in it.
First off we can use Moz’s numbers but thats not scientific.
Broken up by industry:
Moz Claims The Average DA is
Thats bull shit. Do government sites fall into the list? Where does Snapchat fall? How is it that the average Health DA is 67 and the largest hospital network in my state gets a 52? Healthcare is huge in my area … the first heart transplant and the first hand transplant came from here. But the largest provider is below average?
If there is an average here … does that seriously attempt to suggest that about half of the sites online are above DA60 ? Cause thats extremely high for any actual company that I’ve looked into.
No these numbers are just way inflated.
Creating A Sample Size For Average DA and CF, TF
I’ll go ahead and make some numbers here using a wide range of hundreds of domains. You can see the values below, check the methodology of the sample in the accordions.
Open the accordion to learn how I came up with what I’ll call our sample of websites.
So lets try another way thats semi random but provides us websites. First I’ll put in websites for everything on my desk. I didn’t plan to look up usedcardboardboxes.com but I have a marker here on my desk so it goes in. That will give us several dozens sites.
We’ll also toss in 25 randomly picked domains from the list of sites that expired in the last 30 days.
But lets get another sample … the employment websites of the first couple pages of my linkedin friends. I didn’t chose them for that purpose so its semi-random. These are some established sites so lets now pull sites that are different.
We’ll add 11 Wix site addresses from an article. I’ll also include the sites on wikipedia.org as most popular websites. Those are not exactly representational but I’ll add the 80 here.
I’ll also find similar sites for each of these sites based on Similarweb.
And then a stroke of brillance. Do you know what the most common anchor text is to sites? “Read More” so I’m pulling a list of domains that rank for that Google search…there are 25 billion urls here so thats pull a swath of unique domains.
I also wanted to ensure a wide net and figured Id pull backlinks from cnn.com in all that was 400 at random. From the UK, EU or Australia. Wanted to ensure it included outside the US
In total that provided 756 domains, I felt this was probably good..
In the end I have about 756 domains here. This is a good sampling in my mind others may disagree. For now it will be the source of our data and I’m offering something really cool…comparing metrics.
You can check out the raw data pulled in my report at the link below: https://docs.google.com/spreadsheets/d/1kCNt4gRCNjPsQ2mMUOGhT6z_cLE_BDSa3LyGKE1H4MA/
So the results…
Average DA CF TF
Note that the AS only includes 528 domains. I didn’t have the ability to pull the data on the added 200 or so I put in. BUT thats a large amount of the same domains.
Lets visualize the data we’ve produced:
DA CF TF AS Compared
Keep in mind that these are from 3 different sources and are meant to measure somewhat slightly different but comparable factors.
Generally the results suggest that an average site is still in the 20s or 30s of any metric. AS may be the hardest to gain value from.
To compare DA to CF to TF to AS we’ll look at an average of the 528 domains numbers and then pull an average of those.
Now just to see what happens. We’ll toss out the domain if it gets a 0 in any metric. Then provide averages of that sample.
This left us with 199 domains with these averages.
Any DA CF TF or AS at about a 30 to 20 is likely average. TF or Trust Flow is the harder of the metrics. 20 to 15 can be considered average.
All the numbers average out in a range of about 20 points of each other, that can be huge so I still would use more than one value.
Difference Between DA and AS
Also interestingly the avg difference between DA and AS is 2. Now they don’t mirror each other that is the average. The range was 26 to 0 on the difference of each domain.
What to do with a low ranking site?
I was asked a seemingly simple question by a client today. Should low level sites he owns be scrapped and used for their backlinks by 301 redirecting the sites.
Simplest Answer: I think it’s hard to quantify if a site is low level.
Ultimateseo.org has a DA 19, that seems pretty low. But keep in mind it used to be a DA 40 before Moz changed it’s scoring. Keep also in mind that change in scoring showed no reflection in the site’s ranking. I recommend 301 redirecting based on anchor text not necessarily on the score a site earns.
Longer Answer: What backlinks should be 301 redirected from another domain?
While it is a Domain Authority 19 now it is a CF 34, and a TF 14. It has a DR 45, and a DS 24, TS 22, AS 36 ….. head spinning yet? As a reminder here are what those abbreviations mean … and to whom.
- DA is Domain Authority created by Moz
- CF is Citation Flow created by Majestic
- TF is Trust Flow created by Majestic
- DR is Domain Rank created by AHREFS
- DS is Domain Score created by SEMRush
- TS is Trust Score by SEMRush
- AS is Authority Score by SEMRush
DA is Unreliable To Me
The DA metric lost a lot of credence in my mind when they redid their formula, it made it more obvious that it’s just a number in a vacuum. What we saw after Moz released DA2.0 was as a DA40 became a DA19 it wasn’t reflective by what Google saw.
The day before this site ranked as it did the day after DA2.0. Google doesn’t use DA and what we care about is what does Google think about a site.
So Id say if we want to redirect 301 links we can …. and I do. But I do it because the anchor text is what I want, not because a site’s DA is low.
Assessing A Site’s SEO Value
When we assess if a site is high or low we have to consider all of the SEO metrics we have available and use these numbers together. Keep in mind they are not intended to be apples to apples. For instance Citation Flow is how influential a site appears where Trust Flow and Trust Score are meant to tell us how trustworthy a site is considered. AS is Authority Score which I lump in with CF. I think of DA, DR and DS as comparable metrics.
So assessing if this site is low or high I’d say we have the following metrics:
Overall scores: DA: 19 DR: 45 DS: 24
Trustworthiness: TF: 14 TS: 22
Authority: AS: 36 CF: 34
It’s also important to look at some of the raw indicators to understand the values we are seeing above. I use Majestic, SEMRush And Moz to answer these questions, and I always expect Majestic to have higher numbers because it crawls more of the web than Moz.
Moz likely would say they crawl more of the relevant web … but who are they to decide whats relevant and I dont care what they think is relevant … just what Google crawls.
Backlinks: So how many links to my site are there?
Backlinks . Moz: 18445 Majestic: 139903 SEMRush 1000000*
Backlinks are fine and dandy but keep in mind a million from one website vs a hundred from a hundred sites says something. So what is the “Domain Pop” … how many domains link to our site?
Domain Pop: MAJ 1302 SEMRush: 4,000
Domains can be owned and operated easily from the same cheap shared host. So how many seperate IP Addresses link to our site?
IP Pop: SEMRush 885 Majestic 905
These numbers above 885 and 905 seem a lot more comparable to each other than some of the other numbers.
Now consider a site that may have a DA: 50 vs Ultimate SEO’s DA: 19. I’m willing to bet 9 times out of 10 Ultimate SEO will out rank the other site as long as the other site has 500 or less domains on different IP addresses.
Better Backlinks and Best Backlinks
Now to be fair this explanation above is overly simplifying the answer. You may wonder how that could be … it seems rather complicated. Well ask yourself which site would you rank higher … a site that has a link to it from 20 research universities or a site that has 200 links to it from random weirdly named domains? Obviously the 20 research universities suggests a more authoritative and trustworthy site.
Authority On The Topic
But getting more complicated what if the page we are considering for ranking is about 2000 dance mixes. The 20 research universities are all medical schools and the anchor text reads “open heart surgery mortality rates” … do we think that the site should rank that high now…and how authoritative do we think the site is on dance remixes of 2000?
301 Backlinks Based On Backlinks Not Your Domain
So the best answer I can give for when to 301 redirect a domain or not is to ignore any one metric completely. If multiple numbers lead you to believe a site isn’t going to be good then sure consider each backlink individually. Keeping in mind what earned the site a low metric across the board IS mostly those backlinks.
If you find a .gov or .edu link using the anchor text you are after on another site then yes … thats the time to 301 redirect that backlink to the best content you have that is relevant to that anchor text. That last part is essential.
You can’t expect long term gains by redirecting a backlink to content that isn’t able to hold its own. User behavior is a big factor in Google’s ranking of a site. And it should be, you can’t prevent what you cant predict.
After all is said and done if no one stays on a site for more than 10 seconds or if they always click the result after yours in Google’s search results you must ask yourself why users hate your content.
And I mean it … people hate your content if you have 301 redirected, optimized your on page and studied your backlinks but users still go elsewhere. Google may rank you high for a time but without users liking your content and staying for awhile you will drop in rankings due to user behavior.