Penguin 4.0 update now live
This is the first time Google has updated the algorithm since Penguin 3.0 in October 2014.
The algorithm is “now more granular,” meaning that affected sites will disappear for certain keywords (or have certain pages disappear – we’ll have to wait and see which) rather than the whole website falling off the face of the earth, like this:
Penguin is now real-time. Illyes says
Historically, the list of sites affected by Penguin was periodically refreshed at the same time. Once a webmaster considerably improved their site and its presence on the internet, many of Google’s algorithms would take that into consideration very fast, but others, like Penguin, needed to be refreshed. With this change, Penguin’s data is refreshed in real time, so changes will be visible much faster, typically taking effect shortly after we recrawl and reindex a page. It also means we’re not going to comment on future refreshes.
This means several things.
Pretty much everyone should see some kind of recovery
Brands tend to “over-optimise” for certain keywords, which is likely what caused them to be affected by Penguin in the first place. This tends to refer to lots of anchor text links, especially links pointing to internal “commercial” pages – loads of “car insurance” anchor text to your /car-insurance page, for example.
Previous iterations of Penguin would effectively trash your whole website. This “more granular” version is likely to let other pages that don’t display the same spam signals return (your /life-insurance page, perhaps).
You can also probably recover a keyword by 404ing an affected page and putting a new one up without redirects. I’ll be testing that this week and will update as soon as we have results because…
It’s going to be weeks before we see the full impact
Let’s assume that core-algorithm Penguin runs in the same way as core-algorithm Panda, which seems to roll out every two weeks and take the full two weeks to roll out.
Those waiting to recover from Penguin shouldn’t get too disappointed if nothing has happened by Monday. It’ll probably be the first week of October before the dust starts to settle and we know who’s been affected and who’s recovered.
We’ll never hear about Penguin again
Google won’t mention Penguin again, except to deny its existence (although this update will give some SEOs a year’s worth of conference material).
— Gary Illyes (@methode) January 19, 2016
In future, the pattern will be as follows:
- Some sites won’t recover right now – because their owners haven’t put in the necessary work, perhaps – or some sites will be hit in future. It’ll be a few years before there are no sites left waiting for a Penguin recovery.
- SEOs will be monitoring the traffic of these sites (and rankings for as long as Google continues to let us), see a massive spike that could only mean a Penguin recovery, and head straight to the forums/Twitter.
- Google will deny a Penguin update within a few days.
- Several weeks later Google will reluctantly admit to updating the “core algorithm” – which Penguin is now part of.
Since Panda was baked into the core algorithm in January there have been several obvious algorithm changes (including in January itself) that saw big changes for sites affected by that algorithm.
I’m assuming that “baking” an algorithm into the core, apart from sounding delicious, effectively means replacing the previous algorithm with new infrastructure. I assume that Hummingbird replaced Caffeine in roughly the same way – it’s a newer, better way of doing roughly the same thing.
In the coming months we’re going to see days when an update has obviously taken place and certain sites previously affected by Penguin recover together – followed by an announcement that a core algorithm update has taken place.
This will signal what we used to call a Penguin “update”.
But we should never have to wait for a “refresh” again.
- Dr. Pete has already posted some great analysis of the run-up to the current update (including that recent “core update” which looked a whole lot like Penguin).
- Martin Woods posted some early thoughts on the update which I largely agree with (including contributions from Andrew Girdwood and Gerry White).