How to Evaluate the Quality of a Link?
Link evaluation is about setting priorities for links you want to pursue rather than ranking them. Here’s how to properly evaluate the quality of a link in relation to your site.
I have to admit something.
I used to buy links.
I used to buy a lot of links.
In my career, I’d estimate I’ve purchased around $3 million in links.
But I haven’t bought a single link in five years. Well, let’s just say I haven’t directly bought a link in five years.
Buying links today is super risky – no matter what the seller may say.
Yes, there are ways to still effectively buy links, but the bottom line is that if you get caught, and you have any type of brand at all, the risk outweighs the benefit.
It’s up to you if you want to purchase links, but my recommendation is to stay clean.
Building links takes more work and more time to achieve results, but the results are solid and will last.
One benefit of buying so many links over the years is that I have learned how to evaluate what a link is worth.
In the next few sections, I’m going to show you how we evaluate links and give our reasoning behind the scoring.
First, it is important to understand that specific links are valued differently by every site.
A link that is valuable to a plumber is not necessarily valuable to a Payday loan company, and vice-versa. That’s actually one of the downfalls of buying links – besides the whole penalty thing.
Typically, when you buy links, you are forced to buy something that isn’t as relevant to your site as you would like – to use the cliche, a square peg in a round hole.
That’s why link building must be custom for every site. No, this approach isn’t scalable in the traditional sense, but it drives the needle in SEO results.
Bottom line: Links that provide value beyond SEO are the most valuable links to obtain.
It has been said that SEO is both an art and a science. This is very true when it comes to evaluating links.
To properly evaluate the quality of a link in relation to your site, you have to make some educated guesses.
Therefore, link evaluation is more about setting priorities for the links you want to pursue rather than ranking them or creating an equivalent ad value or monetary equivalent.
This process is designed to help you know where to focus next.
The Link Evaluation Process: Getting Started
The Perception score is the qualitative data in our formula. It’s data that comes from a human, not a toolset.
Take your list of sites and influencers and evaluate them based on how relevant they are to your current outreach. We use a scale of 1-100, with 100 being the perfect link and 1 being not relevant at all.
For example, if you are running a site selling online training to insurance sales professionals, a link from a site that outlines the latest insurance training techniques would score very high – taking into consideration whether the site is a competitor.
On the other hand, a site that focuses on fly fishing might contain some of your audience, but a link there will not be as relevant. Notice I didn’t say valuable.
The fly fishing site may be of value – but it’s not as relevant as the insurance site.
Relevance is part of the formula, but not the whole. And relevancy does not always equal value.
We want to assign how difficult we anticipate it will be to obtain a link on a site. As you build links, your experience will help guide you on this metric – and it’s never perfect.
In fact, I recommend you change this metric on your list as you learn more about the site. This is an ongoing evaluation – not a one-time score.
For this metric, we use a scale of 1-50, with 50 being a link that extremely easy to obtain (think a link you can place yourself) and a publication like The Wall Street Journal being at the bottom end of the scale.
Some items to consider when scoring the difficulty of a site include:
- How often the site publishes.
- The number of visitors a site receives each month (be careful here though, some sites receive lots of traffic to certain sections and no traffic to others).
- Relationships with your competitors.
- The use of nofollow links on the site.
- The propensity of the influencers that write for the site to engage with their audience.
This metric is hard to get right – in fact, you’ll probably never nail it right on the head. But getting in the ballpark is usually all you need to do.
The Relevancy Score and Difficulty score combine to create the qualitative portion of our evaluation. We call this the perception score.
If you are concerned that your perception score is either biased or otherwise incorrect, you can utilize intercoder reliability.
Intercoder reliability is a fancy way of saying use three people to do the evaluation.
By taking three scores and averaging them, you are more likely to come up with a number that is in the ballpark of accurate.
Hard Number Score
Back in the day, sites were priced almost exclusively on their PageRank – a number we all saw in a little green bar that Google provided to us in their toolbar. This was highly inaccurate data and really didn’t tell us much, but that’s how the marketplace worked.
PageRank was a “hard number” that was the agreed upon metric for link evaluation. With the proliferation of tools like Moz, Majestic, Ahrefs, and others, we actually have more hard numbers than we did back then.
We use Moz’s metrics, specifically the metrics in Moz’s Open Site Explorer, for our evaluations.
Note: some of these metrics require a subscription to Moz, which I highly recommend to anyone in SEO. You can get similar data from the other tools, so use that you prefer.
But to get the right data, you’re probably going to need to sign up for a subscription. You get what you pay for.
The first metric in the hard score that we look at is Domain Authority (DA). This can be a misleading metric for several sites, especially if the site has sections that are more popular than others.
If there are specific sections of a site where you envision your links appearing, you will need to look at the Page Authority (PA) of the pages in the section – but for most sites, DA, while an imperfect metric, is fine.
If you do need to use PA, it may mess up the formula a bit – but usually not enough to matter. Just make sure to pay close attention to any site where you are using PA instead of DA.
Next, we look at the established links metrics. Basically, we are looking to see how many links a site has.
The number of root domains is typically the most valuable metric when evaluating a link.
I typically only use the Total Links metric as a tiebreaker.
Identifying Spam Sites
Look closely at the difference between the Root Domains and the Total Links. A large differential can indicate that the site may have some SEO issues.
Looking at this differential alongside Moz’s Spam Analysis can help you avoid targeting links that may harm your site.
Every client’s risk tolerance is different, so there isn’t a magic number that tells you if a site is spammy – but if you suspect it might be spam, probably best to avoid it.
Weighing the Metrics
Now it’s time to put all of our data together and create evaluation scores. To do this, we need to weight each metric in order to come up with our final score.
There is no “wrong” way to weigh these metrics, as long as the end score helps you to prioritize which links to spend resources obtaining first.
First, you need to take the sites and compare the Root Domains, ranking them highest to lowest according to the Root Domains metric.
Then divide the sites into groups of 10. If you are evaluating less than 10 sites, each site will be its own group.
Label the groups from 1-10, with 10 being the sites with the highest number of links and 1 being the group with the lowest number of links.
If you have less than 10 total sites, label your highest site as a 10 and then count down from there. We call this the Root Domain Metric.
Next, take the Domain Authority of each site and add the domain authority number and the root domain metric together and then divide by 2. The reason we divide by 2 is that we feel that these metrics are half as indicative of the value of a site vs. the qualitative metrics. This number is called the “hard number score”.
Now we move on to the qualitative, or as we call it, the Perception Number score. This is simple to do. Just add the Relevancy score and the Difficulty score.
The thinking behind this is that Relevancy is twice as important as difficulty in link evaluation, hence the reason we measure relevance with a scale of 1-100 and difficulty with a scale of 1-50. But also, the easier a link is to obtain, the better – hence the reason we rank easy sites at the higher end of the difficulty scale.
In the end, you’ll add the perception score and the hard number score to get an overall link evaluation score. The higher scoring sites should be the ones first on your list.
|Site||Relevancy||Difficulty||Perception Score||Domain Authority||Root Domain Score||Hard Number Score||Evaluation Score|
Use Your Brain
Obviously, this is not a perfect system. The biggest downfalls usually happen in the “difficulty” stage where easy links are in sections of sites that provide no value.
You have to look at every site. I can’t stress the importance of having a person check out every site to see where your links would fit in.
Also, don’t skip the Spam evaluation step. Getting a link placed on a troublesome site can harm you.
This simple formula should help you prioritize the links you should work to obtain. If this is valuable to you or you have suggestions to make it better, I’d love to hear from you on Twitter.