2012 was an amazing year for link building, a lot changed and to be honest I’m glad those changes happened. I think it is necessary in order for our industry to grow up and be taken more seriously in the marketing space. A lot of us add a huge amount of value to businesses, it’s time we were recognised for that rather than relying on and being remembered for low quality tactics that should never have worked in the first place.
The first step for many is to undo the work previously done and this starts with finding the low quality links that may have been built many years ago. If you’ve been hit by a penalty, then you’ll also probably want to get them removed. This post is a step-by-step guide to finding what are likely to be low quality links.
This post will use a number of tools to help, however you need to use your own judgement and common sense too and not remove links which are clearly not hurting. I heard stories of people emailing Yoast to remove links from his site – that is just crazy.
Step 1 – Grab all of your links and linking domains
There are lots of tools out there which tell you who is linking to you, unfortunately there isn’t a single tool that can tell you all of them. So you need to grab data from every tool and use Excel to make a single list which you can work from. Here is a list of places you can grab your links from:
All of these tools allow you to pull your list of links into a spreadsheet, so I’d start by building a spreadsheet with the following sheets as my basis:
You will need to download two types of links – linking domains and individual links. You can either add more sheets for these or do the analysis for each type in a totally different spreadsheet. To keep things simple for now, I’m going to create a new spreadsheet for my linking domains.
Now we need to aggregate all of these sheets together so that we have a single list of all links and linking domains. This is where it can get a little bit messy because the different tools will have different metrics, for example Open Site Explorer has domain and page authority whilst Majestic has citation and trust flow. To keep things simple, I prefer to choose a few metrics which I use throughout the rest of my analysis. This choice is yours depending on which tools and metrics you prefer, but I like to use the following which form the headings of my new sheet:
As I said, you can use whichever metrics you prefer, the key is to be consistent across all of your links.
Now things can get a little bit messy because we need to start copying over the data from different tools into this single sheet – whilst keeping the metrics we already have and keeping things organised. So a little bit of copy / paste and moving columns around is required. For example I need to grab data from my Majestic download and paste across the values I need. I start by deleting the columns I don’t need:
Again just to reiterate, this is just my personal preference. If you prefer to use these metrics, by all means keep them!
Once I’ve deleted the data I don’t need, it is simply a case of copying and pasting each column into my main sheet, matching up the column headings as I go. Once I’ve finished, there are going to be gaps in the data, for example the links I’ve copied from Majestic will not have domain or page authority metrics. Fortunately there are some great tools out there to easily grab the missing data:
- Links API extension for Excel – can grab link metrics from Open Site Explorer and Majestic
- SEO tools for Excel – can grab tons of stuff, in this case I want to grab the HTTP status code and PageRank
These tools are amazing. It may take some time to pull in the metrics you need, but at the end of this step you should have a sheet that lists every single link you know about along with the key metrics that you want to use to measure their quality.
Step 2 – Find the low PageRank links and visualize
There are a couple of ways of doing this, the first way is to use a simple filter on your PageRank column:
The second way looks a bit more fancy and is good for giving you an idea of the overall picture of your link profile:
Step 3 – Find your sitewide links
Not all site wide links hurt you, they are a normal way of the web if you imagine blogroll links. However the combination of site wide links and a low quality linking domain is a bad one when it comes to SEO. So now we need to find our sitewide links and cross reference them with the quality of the domain. If we find sitewide links from low quality domains, then those should be prioritised in our link removal.
The advantage to prioritising these also lies in the fact that you only need to contact one website owner to get many links taken down at the same time, so it can be a lot more efficient.
To find site wide links from our data, we need another simple pivot table so that we end up with something like this:
The data above comes from the backlink profile of my own blog, so as a sense check I wouldn’t worry about removing links from these sites 🙂 You will need to check the PageRank of these domains and if they are 0 or -1, then they should be prioritised for removal.
Step 4 – Find your over-optimised commercial anchor text
I’ve always struggled with anchor text as a signal, it just isn’t natural at all for a site to have lots of commercial anchor text in their link profile, yet Google value it so highly. Anyway… this is one of the easiest ways Google can deem your website to be over-optimised. As with the previous point on sitewide links, I’d advise a bit of caution because having commercial anchor text isn’t neccessarily a bad thing. But having way too much of it or having it coming from lots of low quality websites is.
Finding your most popular anchor text is easy using the standard backlink tools where we got our data from, you simply click the relevant report tab. However remember that we want a clear picture of our entire link profile, so we need to create another pivot table form our main spreadsheet of links. This time, instead of pivoting by linking URL or PageRank, we’re going to pivot by anchor text:
Again, the anchor text pointing at my own site looks pretty good and if this is the case, don’t worry too much. However if your top anchor text is made up of pretty much all commercial anchor text and no branded at all, then you should take a closer look at those links. If they also prove to be low quality, these should be prioritised for removal too.
So in the steps above, we have found:
- Links that are from low quality domains
- Links that are sitewide
- Links that have too much anchor text
All of these alone should give you plenty to focus on and to remove lots of low quality links. Remember to sense check your list first to make sure you’ve got no false positives in your list and accidently removing good links!