Compared to 2014 the amount of major updates to Google’s ranking algorithms was quite small in number, however those that were released had a major impact on Business Owners, Online Marketers and searchers. Here’s an overview of what was implemented in 2015:
April: Mobile Friendly AKA Mobilegeddon
In a somewhat unusual move Google announced this update ahead of time, giving website owners a chance to make their websites “mobile friendly” in order to not lose traffic from searchers using their mobile device. The update is quite simple: if your website is not responsive and doesn’t provide a positive experience for visitors using their mobile device, your website won’t perform as well in the search results.
The algorithm operates on a page by page basis, so if your website isn’t completely mobile friendly but some pages are they could still benefit. There are no different grades of “mobile friendliness” – a page is either mobile friendly or not!
To quickly determine if your web pages are “mobile friendly”, Google released a Mobile Friendly Testing Tool.
May: The Quality Update
In May there were changes to Google’s core ranking algorithm in terms of how it processes quality signals. It was initially believed to have centred around the Panda update, but Google was quick to deny that.
Specifics of how quality was assessed wasn’t released, with Google sticking with the statement of the quality pages being determined by a wide range of individual factors. Some of those factors are now being weighted differently to before.
July: Google Panda 4.2
In July Google announced that the first update in ten months to its Panda Algorithm had been launched. The rollout was slow and website owners that had previously fallen victim to the algorithm did not see an upward turn in their Google traffic overnight.
October: Rank Brain
October saw the news emerge that Google had been using a machine-learning artificial intelligence system called “RankBrain” to help sort through its search results since near the start of 2015.
Google told Bloomberg that a “very large fraction” of queries are being processed by RankBrain. Interestingly in terms of learning, RankBrain does this offline, by being given batches of historical searches and learns to make predictions from these.
If those predictions are tested and if proven good, then the latest version of RankBrain goes live. Then the learn-offline-and-test cycle is repeated.