Everything you need to know about SEO, delivered every Thursday.
Taking A Closer Look At Link Changes
When you’re trying to build as natural a backlink profile as possible, you need to carefully think about all the things that could make it look strange. If your typical link growth each month over the past 12 months has been 5 links, getting 100 will seem odd. Suddenly focusing on an incredibly bizarre long-tailed keyword in your anchor text may signal that you’re using the links to up your rankings.
We keep hearing about all the latest and greatest new way to build links, things we “must” do, but we don’t always step back and consider what a natural link profile actually looks like as we try to beat everyone else in the SERPs.
When doing a backlink audit, there are a few things that automatically make me ask questions, namely “what happened here?” Since I firmly believe that having a good idea of a link history is critical before moving forward with a campaign, and I am slowly but surely becoming less nonchalant about risk, I thought I’d share these items with you and explain why they concern me.
Now, none of this is intended to say that these occurrences are indicators of Google violations, etc. They’re simply changes that make me take a closer look. The more risk-averse I become, the more I don’t like to find anything that makes me take a closer look, if you get my drift.
Link spikes are a natural phenomenon in link building, occurring whenever you get an increased number of links in a small period of time. They can also serve as an indicator that something unnatural is happening with your site’s marketing. If I view my site’s link spikes over the past year, they occur whenever I have an article published.
Over the course of time, they appear in a very natural manner, as they happen at certain increments. I’ve examined backlink history graphs that look very nice and have gradual increases each month, then there’s a massive spike that can be traced to a piece of linkbait.
To me, that seems ok as there’s an explanation for it, but I can see that it would seem odd if I had no idea of what caused the spike. While I’m not as worried about the signals that link spikes send as I once was, I do think that they can look very out of place, and that concerns me. However, since I don’t work for Google, I’ll hush about it now.
Changes in Link Types
Remember all the fuss surrounding no-follow ages ago? We were told that it needed to be placed on paid links so we wouldn’t be penalized. People rushed to no-follow their paid links (and probably some that just looked sketchy)…can you imagine how that changed all the data for a site?
One month, you do a report and you have 50 nofollowed links, then the next month, all of a sudden you have 1000. Weirdness. If I had been watching a site for a few months, I would have wondered why the links all changed to nofollow, and of course I’d be thinking that it was because those were all paid.
(While not a change per se, I also find it very strange to come across a page full of links that are no-followed except for one big glaring and most likely paid link.)
Changes In Link Targets
If you have been around for a few years and your backlink profile is 98% homepage links, it would look odd to immediately get 1000 new subpage ones, right? Similarly, if you’re usually getting a sprinkling of links here and there to subpages and suddenly get 500 to one brand new page, it’s going to look unusual.
Of course, there could be a logical explanation but remember that the explanation may not be obvious. Again, as with all of this, changes in link types can happen in a very natural manner, but a significant switch to subpages from the homepage (or vice versa) can definitely look very strange.
Anchor Text Variety
This is one of the ones that concerns me the most as a link builder. We’ve done audits of profiles that have almost no keywordized links. They are 98% brand/site/URL anchors. These people want to build 500 links to a long-tailed phrase such as “magical unicorn wizard hats that glow in the dark.”
Imagine seeing that profile, as you’d immediately cringe, and not just because of the wizard aspect. It just looks like someone decided he wants to rank for that term for his site, and he’s gone out and (probably) bought a bunch of spammy links because let’s face it, no one is going to naturally link to you with that anchor text.
In general, don’t take a kneejerk reaction to link building, especially without considering how you’ve done things in the past. If you decide that you need to build more brand/URL/keyword/nofollow/image/directory links, then do so by all means, but don’t just stop doing what has been working for you. Algorithms change and new things are constantly being touted as the tactic you need, but it’s just not a good idea to keep drastically switching your methods, either.
I have seen no conclusive evidence that either confirms or denies the idea that you should always keep your link building as natural-looking as possible in order to rank well. We can rank sites with paid links for ultra-competitive terms and those rankings stick. We can lose rankings in exactly the same way.
However, that does not at all mean that attempting risky methods is a good idea. What works for one niche or site may not work for yours. What works today may not work next week. This is why, in the end, it’s probably best to do your utmost to keep your links looking as organic as you possibly can, and your link history plays a big role in exactly how you do that.
Some opinions expressed in this article may be those of a guest author and not necessarily Search Engine Land. Staff authors are listed here.