Ouch. It appears that the Australian insurance site realinsurance.com.au is in back-pedalling defensive mode in removing some dodgy links to their site.
It appears that an SEO firm, no longer working with the company, did a lot of blog commenting to build low-quality but just-about-worked links to the site. This is an oldish tactic that I guess many of us know, and probably don’t love, and a tactic that Google’s recent Penguin update was designed to neutralise.
If you’re unfamiliar with how it works, it’s the act of leaving a comment on someone’s blog, with either the name of the commenter linking through to your website, or a link embedded in the comment itself.
Here’s an example of a comment linking to Real Insurance, found on GetInTheCarReview.blogspot.com:
Can We Have Our Link Back?
High-quality stuff, I’m sure you’ll agree. Anyway, it appears from a report floating around the web that someone on behalf of RealInsurance.com.au has been appointed to contact the webmasters of all the sites featuring comment spam, and requesting its removal.
Here’s part of the justification for the request, as obtained from TPM:
This request is being made because your domain and link has been identified as one that may be causing our site a penalty post Google Penguin update. The work of previous SEO agencies on our behalf has caused problems for us post Panda so I hope you can help us comply with Google guidelines.
Those pesky previous SEO agencies! Oh and, if we’re being pedantic, it’s the Penguin update and not Panda, which was all about content and nothing to do with off-page links.
The Damage Done
So just how bad is the problem?
Well, let’s start with a breakdown on the types of links that are currently linking to Realinsurance.com.au. The below chart is an export from the wonderful LinkDetective.com, and shows that most of the links pointing to Real are mostly low-quality directory links, blogroll links (which are quite often paid), forum posts and a lot of other shoddy-looking links:
This data is based on Open Site Explorer’s discovery of 3,446 links which is somewhat shy of the 103,000 reported by Ahrefs, which I consider to be a far superior tool (but which, unfortunately, LinkDetective doesn’t play nicely with.) We can consider the 3,446 to be a reasonably arbitrary slice of the real number of links.
It’s striking just how many of these links are the sort of low-grade ones that, for all intents and purposes, anyone can get. There’s no barrier to gaining a link by registering a forum profile, no editorial “vote” in a blog comment, and sidebar and blogroll links are typically paid-for inclusions using services like Automatic Backlinks.
All in all, it’s not a good look for an industry-leading insurance company.
Can I Buy Anchor Insurance?
LinkDetective will also show us the weighting of anchor text among these links; which more or less equates to an overview on the keywords Real wanted to rank for:
Here’s some similar data from Ahrefs.com showing just how many links are out there with individual anchor texts:
Again, Ahrefs uses a different (superior, more reliable) dataset than Open Site Explorer, and I apologise for the mismatch of numbers here. What’s interesting though is just how many links there are for the top phrase seniors insurance – over 55,000 of them, across just 23 different domains. Ouch.
Erasing The Past
Ahrefs gives a pretty neat overview on backlinks gained and lost, which shows that a fair few sites appear to have complied with the link takedown request:
There’s a mismash of sites showing up as lost backlinks, but most of them appear to be within the web directory and blog space. A couple of hundred links might have been removed over June, but that’s a minor dent in over 100k links to fix up.
RealInsurance – Good Or Bad?
So, the industry has noted what they’re doing, and has done a little light analysis on how bad the problem is. And, I guess, we’re going to conclude that it’s quite a bad problem.
Clearly they should never have got themselves into this mess in the first place, but claiming hindsight isn’t going to be much help. Insurance is a notoriously competitive niche, often held up as an example of how competitive a niche can be. Real might have found themselves in a situation where everyone else was doing spammy SEO, and had to follow suit in order to stay competitive.
(And let’s be honest, as SEOs I’m sure we’ve all had conversations with clients around how they’re supposed to compete against dodgier sites by staying completely straight-and-narrow; an easier conversation to have since Penguin, but it was a different story two years ago.)
So are Real doing the right thing by asking for links to be taken down? In theory I’d say the answer’s yes: there’s not a lot else you can do when faced with fixing up bad links, at least until such time as Google launches their own version of Bing’s Disavow Links tool, which lets webmasters manually “turn off” inbound links.
Where Real are exposing themselves to risk lies in how they’re going about the problem. Most bloggers care little for SEO, and care even less for blog comment spam. Asking a blog owner to take time out of their schedule to jump into archives, find articles with the right comment links, and remove them is a hell of an ask — why wouldn’t a blogger spend the time creating a post instead, especially now Real has given them something to post about?
It also raises the point, why doesn’t Real Insurance go about creating some new, valuable links instead? If it’s the case that the links they’re trying to remove are the result of low-quality bargain basement SEO work, then why try and treat the problem with the same level of approach?
This is what I’d be saying to Real Insurance if they were listening. Asking for the removal of links is provocation, particularly when dressed up as a cease-and-desist, as some have been alleging. Doing so makes the link removal become a story in and of itself, and this in turn only opens up the possibility for a competitor to build another 5,000 blog comment links to your site using spam techniques, and you’re even worse off than when you started.