UPDATE (16 March 2016): As reported by SEMPost, Google plans to increase the ranking effect of their mobile-friendly algorithm update, probably some time in May 2016. Our study below shows how much impact the initial update (in May 2015) had, especially on non-mobile-friendly pages. Google has said the new update will use the same criteria and filters as the initial update, but give them more “juice” as an overall ranking factor. Read Jennifer Slegg’s report at the link above for more details. We plan to rerun this study after the May 2016 update to measure how much it changes the search rankings.
In this study, we look at the impact of the so called “Mobilegeddon” update. This update was viewed as being a potentially very large update. At SMX Munich, Google’s Zineb Ait Bahajji was quoted as saying that its impact would be larger than Panda or Penguin. In addition, at SMX Advanced, yesterday, June 2nd, Google’s Gary Illyes reiterated that the Mobile Update indeed had a bigger impact than Penguin and Panda combined.
However, media reports since the update have suggested that it was a lot tamer than expected Some have even called it a non-event. So what’s the real story?
Luckily, at STC we set up a test to measure the update’s impact. We pulled ranking data on the top 10 results for 15,235 search queries the week of April 17th (before the algo rollout), and again the week of May 18th. We pulled ranking information, and also identified whether or not the URLs in the results were designated as Mobile Friendly by Google or not.
Note that in the meantime, Google also did their Quality Update, so this will have some bearing on the data as well. However, the numbers below still tell a story that suggests that the impact of the April 21 mobile algo update by Google was bigger than what most people currently believe.
What Happened to the Top 10 URLs as of April 17th?
This is a unique view of the data that we took that tells the story. Basically, we took the URLs that ranked in the top 10 for our test queries, and we saw where they ended up in our May 18th view of the data. In other words, if a URL was in position 4 on April 17th, but showed up in position 54 on May 18th, we tracked that in detail.
Here is what we found:
The original set of Non-Mobile Friendly URLs got nailed. Nearly half of them dropped in rankings, and 2.3 times as many dropped as went up. In contrast, the Mobile Friendly pages fared much better, with an overall increase in rankings.
You might ask: Wait, why did the average rankings for the Mobile Friendly pages not go up more? Great question. There are two reasons why this actually makes a great deal of sense:
- URLs that were in the #1 spot (15,235 of them) had no opportunity to gain in rankings. They could only go down, and 70% of these were Mobile Friendly.
- The quality update from Google caused some shifts in rankings too, and some of the demoted domains from that update were Mobile Friendly.
In conclusion, this view of the data shows that Google did provide a material preference to the original set of Mobile Friendly URLs.
Percent of Mobile Friendly URLs in the Top 10 Results
In addition to seeing what happened to our original top 10 URLs, we also pulled data on the makeup of the current top 10 as of May 18th. The idea was to see if the percentage of the results that are Mobile Friendly has increased. Here is what we found:
- On April 17th, we had 56,164 Non-Mobile Friendly URLs (36.9%), and as of our May 18th measurement, that dropped to 54,162 URLs (35.6%).
- On April 17th, we had 96,186 Mobile Friendly URLs (63.1%), and by May 18th that had increased to 98,188 URLs (64.6%).
This method of testing the results of the April 21 mobile algo release appears to provide a conflicting view to our other slice at the data. The net shift here is only a 1.3% increase of Mobile Friendly URLs in the top 10 results. This view suggests that Mobilegeddon didn’t come close to the impact of Panda or Penguin.
These two data views appear to contradict one another, but given the long slow rollout of the algo, and the release of the Quality Update in the middle, there are many additional factors in play. From the initial URL set, Non-Mobile Friendly URLs saw a significant negative impact, but in many cases, they ended up being replaced by new Non-Mobile Friendly URLs.
Here is a breakdown of what went on:
These other changes are likely the result of 3 factors:
- The Search Quality update
- Other algo tweaks along the way
- General churn that takes place in Google’s search results
In summary, I’d suggest that the impact of this release was indeed significantly bigger than originally met the eye. The trade press did not see it as large because of the slow roll out, and the intervening Search Quality Update.
Update:New research released by Adobe shows that they saw a drop of 12% in traffic to sites that were non-Mobile Friendly. An article released by Alistair Barr of the Wall Street Journal featured both our study (this post) and the Adobe data. You can see that article: Google’s Mobilegeddon was a big deal after all here.
In addition, this is likely just the start of what Google plans to do with this algorithm. It is typical for Google to test some things and see how they work. Once they have tuned it, and gain confidence on how the algo works on the entire web as a data set, they can turn up the volume and make the impact significantly higher.
It’s my expectation that they will do that. In the long run, don’t be surprised if the impact of this algorithm becomes even greater, and that people will stop debating whether or not it was greater than Panda or Penguin.