How Accurate Are The Leading SEO Tools?
We all know links are the Holy Grail when it comes to SEO and ranking well.
With PR (Page Rank) rarely updated we have only two things we can rely on when judging the quality of a link or website, a keen eye and SEO tools. A keen eye and the ability to find a black cat in a coal cellar comes with experience.
SEO tools on the other hand are readily available and can tell you a vast amount of data very quickly with minimal effort. But how accurate are these tools and which measures are the most accurate? Today we’re going to find out.
Chosen niche: Yoga
Tools we’re going to analyse: Majestic SEO, Moz, Ahrefs plus many more
Test synopsis: I want to establish what if any consistencies and trends can be found from the collection of SEO tools available and establish those that are the best measures of a quality link or website. I have taken the top 200 Google UK search results for an un-personalised search of “yoga”. I removed the inevitable news websites, authority sites such as Facebook and Twitter and any other unrelated websites. This left 89 yoga websites.
Page Rank (PR)
With all 89 websites in an excel spreadsheet, the first piece of data I wanted to gather was the Page Rank of each websites homepage. I did this using the SEO plugin for Excel by Niels Bosma. If you haven’t already installed this Excel extension I highly recommend it.
Every single one of the 89 websites checked had at least a page rank of one, not surprising when you consider these are many of the top 200 ranking websites for “yoga” a 40,000+ exact match UK search.
The exact Page Rank figures stack up like this.
|Page Rank||Number of websites|
That in itself is very interesting, so few PR 4 or above websites in the top 200 Google results for such a competitive term is the first of many key take-away facts from this experiment. Also only one PR 1 listing in the whole of our list is really surprising in my opinion. I was expecting a more even split between PR 1 and 2.
Moving on and next I wanted to check which of the websites had a DMOZ listing. If you’re not familiar with DMOZ, it is widely agreed to be one of the most authoritative web directory online and most SEO’s will try at least once to get their clients listed on there.
Only ten of the 89 websites checked had a listing on DMOZ, of these four were listed twice and the other six only once.
The age of a domain and its effects on ranking and performance has been a hotly contested issue for a long time so I thought it would be a great thing to add into our experiment. I checked every one of the 89 websites domain age and found the oldest domain name was fifteen and a half years old with the newest being a little over one and half years old.
|Domain Age||Number of websites|
|5 – 9 years||31|
|3 – 5 years||18|
|0 – 3 years||16|
Domain Authority or DA is a Moz metric that “represents Moz’s best prediction for how a website will perform in search engine rankings”. A Domain Authority check of a website is something that can be done quickly and give you a snap shot of how authoritative a website is. Though I would add that I rarely trust the DA it can be useful when glancing at a website as it takes things like linking root domains into account so can give you an idea of whether a site is established or not.
SEO Tip: Download the Moz Bar for instant DA analysis of any website instantly.
What is a good Domain Authority is something up for debate, I can tell you for me personally I am looking for websites with at least a DA of 20-25 plus when glancing at a website for the first time. Though I do not always adhere to this, I generally found a website with a domain authority of less than twenty is either going to be low quality, a new site or have minimal content. From an SEO standpoint all of these factors would make me think twice when evaluating the link potential.
Of our 89 websites, the Domain Authority is broken down as follows:
|Domain Authority (DA)||Number of websites|
|40 – 49||5|
|30 – 39||31|
|20 – 29||44|
|10 – 19||9|
|0 – 9||0|
What is really interesting now is when we compare the top ten websites from a Domain Authority standpoint with the Page Rank of those same ten websites:
|Domain Authority (DA)||Page Rank (PR)|
It is worth remembering at this point that Page Rank (PR) is a Google metric so to a certain extent we can take to be true and accurate in as far as how authoritative Google believes a website is. What is key here is how accurate Moz’s Domain Authority is. Within the top ten results Moz’s Domain Authority measure has picked up our only PR 5 site and four of the six PR 4 websites from our list of 89.
Another Moz Metric, Page Authority calculates how well a given web page is likely to rank in Google’s search results. The Page Authority measure also features in the Moz Bar alongside Domain Authority. Seeing as Moz’s Domain Authority appeared to perform so well let’s see how well Page Authority stacks up. It is worth remembering that PA is measured much in the same way as DA with a score between 1 and 100.
|Page Authority (PA)||Number of websites|
|40 – 49||23|
|30 – 39||47|
|20 – 29||11|
|10 – 19||2|
|0 – 9||5|
Let’s consider for a moment that the homepage is more often than not the most authoritative page on a website it is surprising to me to see over 1/5 of the total websites checked with a homepage PA under twenty. As we can see the majority of websites have a PA of 30-40 as we would expect, with only one website featuring a homepage PA of 50+.
Taking a closer look at that one 50+ PA website, we can see:
- It has a domain age of 10+ years
- Has a Page Rank (PR) of 4
- Is one of only 5 websites to have a Domain Authority of 40+
- Doesn’t have a DMOZ listing
Other interesting facts from the PA check of the 10 highest ranked sites include:
- All sites have a PR of between 3–5
- All sites have a DA of at least 35+
It looks like another tick in the box for Moz on their Page Authority check.
Let’s move on as things are starting to get interesting. Next we’re going to evaluate what effect the number of linking root domains (LRD’s) have on metrics and their accuracy. Linking root domains are the number of different websites linking to a given site. One website can be linking to a site from every page such as in the footer of sidebar of a site and this is referred to as a “site wide link”. Regardless of the type of link, all we and ultimately Google are interested in is the number of different websites linking to each of our sites we’re experimenting with.
|Referring domains||Number of websites|
|500 – 999||1|
|300 – 499||3|
|200 – 299||9|
|100 – 199||16|
|50 – 99||21|
|0 – 49||38|
It is worth mentioning at this point that the number of referring domains to each of the websites was taken from Majestic SEO. Therefore we can almost guarantee that these links are not all of each websites backlink profile, however it does give us a good sense of the numbers we’re playing with and from my personal experience I have always found Majestic SEO to be reasonably accurate. For one of my websites I know I have around 180 referring domains and Majestic has recorded around 150 of them so that should give you an idea of how extensive their backlink archive is.
This table has surprised me the most so far in this experiment. If we had ran this same test several years ago you could almost have guaranteed each site would have had hundreds if not thousands of referring domains, why? Because almost everyone was spamming in one way or the other.
Looking at that data it is clear to me that some of our 89 websites will have actively removed backlinks while others might have rebranded or simply not attempting link building at all. Only five of our 89 websites have over 300 referring domains.
Of those five are there any clear correlations that jump out? No not really. We have PR ranging from 2-4, with three of the five sites being ten years plus and Domain Authority ranging from 32-46.
So what does this tell us? The number of referring domains does not count for much alone. This conclusion ties in with the statements the SEO tools and Google have been making that it is about quality and relevance over being simply a numbers game. We will get more into that later on for now let’s make a note that the number referring domains alone is not going to make a huge difference to our SEO efforts and move on.
Links from authority websites such as those owned and operated by Governments, schools, Universities and alike are generally seen as some of the best links you can gain for your website. Typically these domains have extensions ending with .gov.uk, .ac.uk or .edu. While we’re investigating referring domains it makes sense to look at the how these types of links fit into the mix.
|Authority Domains (.gov.uk, .ac.uk etc)||Number of websites|
As we might expect, the vast majority of our 89 websites do not show any .Gov.UK or similar links. However the fact that 21 of the 89 sites do have at least one referring domain from these authority sites shows some high quality link building is going on in one form or another.
There is a clear winner, a yoga website that has six referring domains from these quality sites. Let’s dig a little deeper into that site and see if any of the other metrics reflect such a significant number of authority links compared with the competition.
Here are the vitals for that domain;
- The website has a Page Rank (PR) of three
- No DMOZ listing
- Domain Authority of 34
- Domain age under 5 years
- Homepage Page Authority (PA) of 45
- Around 150 referring domains so average when compared with the competition
The key take-away from that data is that with quality links domain age counts for very little. It further reflects our earlier discovery that SEO link building is no longer a numbers game and instead a quality and relevance game.
What is also really interesting is when we look at the other 16 websites that have at least one authority referring domain, all of them have relative few overall links. None have anywhere near the top numbers of total referring domains we saw earlier.
This is a Majestic metric that measures the quality of the link/website by evaluating all the links to it and cross-referencing with the Majestic’s list of “trusted websites”. A simple way of thinking about Trust Flow is that it looks to see the neighbourhood your website lives in, if it is surrounded by good quality and relevant sites then you’re in a good neighbourhood and the Trust Flow will reflect this. If the website has a lot of spam and low quality websites, Trust Flow will believe your website lives in a bad neighbourhood and assign a Trust Flow accordingly.
Trust Flow changes upon every Majestic SEO update and recording this data alone and how it changes is very interesting – possibly an experiment for a later date!
Trust Flow is a metric I use almost daily as I find it to be very accurate in establishing a websites “credibility”, let’s see if my faith to this point has been justified.
|Majestic Trust Flow (TF)||Number of websites|
|40 – 49||0|
|30 – 39||6|
|20 – 29||17|
|10 – 19||48|
|0 – 9||18|
Looking specifically at the top ten ranked sites according to Trust Flow (TF), the following stands out;
- Seven out of ten have a DMOZ listing
- Six out of ten have a domain age of 10+ years (all ten are over five years old)
- Referring domains are relatively low in numbers for all ten sites
- Page Rank (PR) has three sites with PR 2, four on PR 3 and three with PR 4
Now what is really interesting is looking at the bottom of the list, the lowest ten websites according to Trust Flow (TF) also have the following;
- No DMOZ listings
- All have fewer than 50 referring domains
- Zero .Gov links
- All have a Domain Authority of less than 25
This data reflects what you might think and you would be right, however what is key to remember is we’re cross-referencing the different data between these SEO tools and so going by Trust Flow (TF) would seem to correlate well with both Moz and what Google is telling us about the websites.
Ahrefs Rank (URL Rank)
Ahrefs Rank (AR) measures the quality and number of DoFollow links from one website to another and takes an average of all those links then calculates an Ahref Rank (AR) for every website checked.
First thing is first; let’s look at how this data breaks down across our experiment sites;
|Ahrefs Rank (URL Rank)||Number of websites|
|30 – 40||4|
|20 – 30||12|
|10 – 20||46|
|0 – 9||27|
The Ahrefs Rank and the Trust Flow from Majestic SEO seem to disagree here. Of the top ten ranked sites according to Ahrefs Rank, only two have a Trust Flow greater than 20 and as I stated earlier 20 is personally the absolute minimum I would generally look for. Does this mean I have been missing out on some quality sites by ignoring those with a Trust Flow in the teens? Let’s find out.
Other interesting Ahrefs data facts include;
- Eight of the sites have a Page Rank of eight or greater meaning Google likes the sites
- Seven of the sites have at least one .Gov domain so we can assume authority sites have some level of impact on the Ahrefs Rank (AR).
- Only two of the ten top sites have a DMOZ listing so maybe not quite so important as we initially thought?
- Nine out of ten of the sites had a Domain Authority (Moz metric) of 30 or greater
- All ten of the sites had a Page Authority (PA) of 35 or greater
This tells me that Ahrefs URL Rank is very closely aligned with the types of results Moz is providing and also very much in line with what Google believes is a quality site. Maybe Ahrefs should be a little higher up our SEO Tools list after all.
The final piece of this rather large experimental puzzle that I wanted to put to the test is social shares, likes and tweets. A great deal is written about their importance so let’s find out first-hand what part they play and how the SEO Tools we have available evaluate and consider these factors when determining the quality of a link/website.
|Number of Facebook Likes||Number of websites|
|500 – 999||10|
|300 – 499||12|
|100 – 299||15|
|0 – 99||50|
I knew social was going to throw up some questions and it hasn’t disappointed. Of the top ten results according to the number of Facebook Likes/Shares;
- Five of the websites have an Ahrefs Rank below twelve
- Eight have a Trust Flow under twenty
- Only one site has over 300 referring domains
- None have DMOZ listings
From this Facebook data I would suggest that Both Ahrefs and Majestic SEO and their corresponding metric tests do not award very much if any direct merit in terms of website/link authority. Whether you feel social signals are important or not, it would seem these tools are not placing a great deal of emphasis on them.
|Number of Twitter Interactions||Number of websites|
|500 – 999||0|
|300 – 499||1|
|100 – 299||1|
|0 – 99||86|
If you look at those numbers you would have to agree the yoga niche doesn’t use Twitter as much as you might think. The vast Majority have very few Twitter shares at all. Quite surprising that 45 of the 89 websites checked had zero Twitter interactions.
Whether it is coincidence or not I do not know but those with Twitter interactions, all be it only a handful of sites, they all were highly authoritative right across the SEO Tool spectrum. Ahrefs, Majestic and Moz all showed some of the most authoritative figures recorded had around 100 or higher Twitter shares… Is there something in that? Do we think if sites in the yoga niche used Twitter on their website a great deal more we would see even greater metric improvements across our SEO Tools?
Across both the Facebook and Twitter tests there were two metrics that showed very strong signs of quality. Domain Authority (DA) and Page Authority (PA) both displayed all ten results as having a greater than average metric which begs the question are Moz more actively looking at social signals as a way of evaluating and establishing quality links/websites. This small test would tend to indicate yes.
Presenting our findings
- Page Rank (PR) is most definitely an indicator of a quality website/link (regardless of the fact it is rarely updated)
- A DMOZ listing will help a website/link authority so list your clients in the most relevant category
- Older domains gain authority more easily and therefore they make great link targets
- High Domain Authority and high Page Rank go hand in hand – Making Moz’s DA check a must have for any SEO
- Page Authority generally comes alongside Domain Authority so increase your DA and your PA will improve as well
- SEO is definitely not a numbers game so stop worrying about creating hundreds of links; instead focus on creating a handful of high quality and relevant links
- Authority referring domains are great to have but exceptionally hard to come by, be creative and see what options there are in your niche
- Pro tip: Do competitor research and filter for .Gov.UK domains to see which authority links your competitors have then attempt to replicate
- Majestic’s Trust Flow seems a little more closely aligned with domain age, DMOZ listings and Domain Authority. I still consider Trust Flow highly accurate and intend to continue using after this experiment
- Ahrefs URL Rank has been somewhat of the dark horse in this experiment, the metric tests showed a close alignment with Google’s Page Rank and Moz’s DA and PA metrics.
- Social is more of a nice to have than essential when it comes to your SEO efforts. The data tests we ran would suggest there is some benefit but it’s marginal at best
If improving your SEO or link building are top priorities on your agenda then I would suggest the following SEO Tools for use, I use all of the following on a daily basis and as you can see from this experiment all of them have their own merit and offer an invaluable insight into what performs well and why.
- The Moz Bar Chrome and FireFox extension
- Majestic SEO (Paid or free subscription options)
- Ahrefs (Paid)
- Excel SEO Plugin (Free)
- Share Tally (Free)
Join the experiment
Did you find some value in this experiment? Have you got any questions or suggestions then get involved in the comments below. I am continually running experiments and would like to stress this is entirely based on the yoga niche. I plan on conducting similar experiments in several other industries to cross-reference the data and look for patterns between niches to get an even better understanding of how our SEO Tools evaluate data so we are best equipped as possible to know which links/websites are going to offer the most benefit.
I hope you enjoyed this niche SEO experiment and if you have any recommendations for the industry I should next focus on let me know in the comments below.