If you’re reading this blog based on the title, I’d assume you’re viewing it with a great deal of cynicism, and rightly so. There’s no such thing as a guaranteed-to-work ‘Penguin Recovery Plan’ because we don’t know the algorithm, we don’t have Google’s data set, and therefore we don’t know crossing which thresholds in which metrics is causing rankings to drop. However, what we do have is the ability to make educated guesses based on data from the sites that we monitor, and identifying patterns in the SERPs belonging to the niches that we target. In this blog, I’ll attempt to do that based on the information that other webmasters have seen (particularly this blog from Microsite Masters), hopefully to produce one of the less-speculative guides to recovering after the ‘Penguin’ update.
One of the main culprits identified that is causing sites to be penalised is an over-optimisation of anchor text. An example of this would be if you targeted the keyword “grey hats” on your main website. Your website’s link profile might consist of 500 links, and 368 of those had “grey hats” as the anchor text, that means 73% of your link profile consists of exact match anchor text. This is very likely to be penalised by Google as it seems like a clear attempt to manipulate your sites link profile to help you rank higher for the term “grey hats”.
To understand this, we have to understand what Google’s definition of a natural link is, that is, a link purposefully created by a webmaster at a website. When webmasters create links, it is highly unlikely that a majority, let alone 73% of them, would create a link pointing to your site with the anchor text exactly matching the main keyword you are targeting. Therefore, Google see this as an unnatural signal, and will penalise for it. Current research seems to show that the threshold for sites in a majority of niches seems to stand at somewhere around 60-65% for exact match anchor text, so if this applies to your site, and you were hit by ‘Penguin’ it might be worth reviewing some of these links.
The key here is to diversify anchor text, but if you’re link building naturally, this shouldn’t be a problem for you anyway.
Another recurring theme seems to be the issue of relevance. In the official Google Blog Post regarding the ‘Penguin’ update, they gave an example featuring a piece of content featuring irrelevant text links. Many people have seen this, and assumed it indicates that Google will penalise, or devalue links that are irrelevant to the context in which they are placed, and to an extent, they are right. Google will penalise for this.
However, instead of becoming fixated by links placed in an irrelevant piece of content, it’s also necessary to think about the domain where the link is placed. A link from a website in a similar, same or related niche is much more likely to be valuable than a link from somewhere completely irrelevant.
Let’s stay with the example of two websites that are both looking to rank for the term “grey hats”. Site A has a link profile consisting of about 200 links, and 128 of them are from domains in the same niche, fashion accessory sites, headwear experts, green hat sites etc. This means that 64% of Site A’s link profile consists of links created on other websites in their niche, a great indicator of authority.
Site B, who are also targeting “grey hats”, have a link profile of about 400 links. However, only 67 of those are from other sites in a similar, or the same niche, this accounts for only 16% of their link profile.
In the ‘Penguin’ update, Site B is much more likely to lose rankings than Site A, and (all things being equal) even if Site B didn’t suffer a ranking loss, site A is still likely to outrank them due to relevance of their links. It’s good to have links coming in from non-relevant websites as well, because a natural link profile will also feature these links (they are perhaps also likely to be the majority, depending on the niche of the site), but they should be reinforced with links from relevant or similar sites where possible.
In this blog I’ve really only focused on the over-optimisation of anchor text from external links, however, there are other signals likely at play in the ‘Penguin’ update. These could possibly include aggressive internal linking with exact-match anchor text, keyword density, bad neighbourhoods, link structures and LSI keyword density (often an indicator of spun content). However, at the moment, I still haven’t seen enough data to form a worthwhile judgement on how these signals are currently affecting rankings. For the time being, it’s best to focus on the one aspect that seems to have had the most impact and it’s a signal that’s relatively simple to harvest data for.
Is Anchor Text a dead metric?
The answer is absolutely not. Anchor text is still an important signal that Google use to determine relevance and authority, and based on Google’s current link graph model, there is no way they could completely switch off its value. What has changed is how they handle it, specifically how they handle those who try to manipulate their rankings through the acquisition of exact match anchor text links.
With the ability of hindsight, it should have been more obvious that this was coming. Looking back at a few of my blogs, I remember writing about some of Google’s Algorithm updates in March, and in a list of 5 that struck me as being particularly significant, two were related to anchor text. Perhaps then, the world should have seen this coming, or at least been more prepared for the way that Google handles anchor text. It’s hard to say exactly whether the changes discussed in the March Search Quality blog where a precursor to ‘Penguin’, but it certainly seems there is a link.
I’d like to sign off this blog by asking for your experiences of the Penguin update, and if you believe its main driver is anchor text? I’d also like to highlight a certain irony that there is in how black hatters are inevitably going manipulate there link profiles further, but just to make them appear more natural. Bearing this in mind, what do you think Google’s next move is going to be?
This blog was written over 6 months ago and Internet Marketing and is an always changing industry which means the information within this blog may be out of date. Use caution when using any methods or suggestions within it.
