In early December, Google announced “With the holidays upon us, it looks like the penguins won’t march until next year.” I’ll be the first to admit that I’m no Tom Hanks, but could this be a clue – the Da Google Code?
Well, regardless of the actual timing, Penguin 4.0 isn’t far away.
There hasn’t been an official Penguin update, for over a year now. The last confirmed update was October 14, 2014. Webmaster trends analyst Gary Illyes was quoted as saying:
“The new Penguin update will make webmaster’s life easier a bit and for most people it will make it a delight.”
If that gives you a warm and fuzzy feeling, think about this; The new algorithm could make life easier for “most” people, but what about the other 49.9%? That leaves a large pool of people who will potentially be unaffected or less than delighted.
In fact, some people will lose their businesses altogether and go bankrupt, if history is any indicator of what lies ahead. Google’s decision to postpone the algorithm rollout is their way of saying Merry Christmas to those folks.
The evolution of Penguin
Before Penguin, bad links were simply devalued and needed to be replaced in order to recover search rankings. After Penguin, bad links became ‘toxic’, requiring a link audit and removal or disavow of spammy links.
Even then, a Penguin refresh was usually required before one could see any signs of recovery. That was a real game changer and the long periods of time, between Penguin refreshes, makes a real-time version appealing to many.
I suspect the new algorithm will take the evolution of link value to the next level – placing a premium on semantically related links. Topical relevance is replacing PageRank. I would expect related editorial links to be more valuable, than ever before.
Google now has reams of data concerning spammy links, thanks to a crowd sourcing program launched at Pubcon back in 2012. You may recognize this program as the Google disavow tool. There’s no question that good links get disavowed every day, but the vast majority of links being disavowed really are spam and the others would be easily identified as outliers.
How semantic search could play a role
The idea behind semantic search is to understand the relationship between things, rather than blindly rely upon strings of data, like meta tags, link anchor text and the like.
The integration of semantics started with Knowledge Graph in May 2012, followed by Hummingbird in September of 2013 and most recently, RankBrain.
Incredibly, machine learning is purportedly the third most important ranking factor, yet its existence was unknown, prior to its disclosure this past October.
The impact of semantic search on the Penguin algorithm could take on many forms. For starters, how about applying RankBrain to an aggregation of all submitted disavow files. In theory, via machine learning, RankBrain could develop a ‘link genome’ for recognizing spammy links allowing for the assignment of a negative value.
On the plus side, RankBrain could be used to measure the topical relevance of a link. This would allow for adding a ‘bonus value’ above and beyond that of traditional PageRank. It would also allow for the devaluation of non-relevant links. If that happens, be prepared for a non-trivial shakeup in the SERPs.
The Death of PBNs
Prior to the first Penguin rollout, there was a wave of manual penalties handed out for unnatural links.
In most cases it was due to the manipulative use of EMAT. ‘Coincidentally’ the Penguin algorithm also punished websites manipulating anchor text. Is it possible that Google tested their algorithm to hand out manual penalties in advance of a wider roll-out?
Now, in December, there is a lot of chatter regarding a spike in manual penalties. Many of those reporting penalties have been using their own Private Blog Networks (PBNs) to boost rankings.
Worse yet, many of these folks are using PBNs for client work. Could this be a test before the full roll-out? If so, PBNs are done and a number of “innocent” clients are going to be affected.
Disavow these kinds of links today, before it’s too late
If you’ve purchased links on Fiverr or have engaged the services of someone who spammed your inbox offering cheap SEO services, you’re due for a visit from Penguin. That would include links like the ones offered by this spammer who emailed me on December 8:
This is just precious: “we use only white hat and ethical SEO techniques for websites.”
I to hate break this to you, but the above links are NOT white hat and they are precisely the kinds of links that we remove ALL of the time to get manual penalties revoked and algorithmic problems resolved.
The semantic web could also lead to the expanded use of unlinked mentions as ranking factors. Google is already including unlinked mentions in the ‘latest links’ file in Google Search Console, leading to speculation that they are already used as a ranking signal. Although John Mu has disputed that saying “they don’t pass PR” as recently as October 15 of this year.