Skip to main content

Research and Studies

Google Still Loves Wikipedia (More Than Its Own Properties)

Particle Wave

In late July, we saw reports that Wikipedia had seen large drops in traffic. This led some to speculate that this Wikipedia Google traffic loss was because Google was favoring its own properties more, at the expense of Wikipedia.
In today’s post, I am going to reveal findings about Google’s results using four different snapshots of rankings data. Here is what the data will show:

  1. The Apparent Cause of Wikipedia’s Traffic Loss
  2. How Wikipedia’s Representation in the SERPs Varies Between Commercial and Informational Queries
  3. A Comparison of Google Properties in the Results vs. Wikipedia

Want quick and easy ways to view and share the results of this study? Jump to the bottom for video and slide deck versions!

TL;DR

Of all the URLs we looked at, nearly 7% of them were from Wikipedia (sometimes more than once per SERP). Looking at it another way, Wikipedia shows up in the top 10 of the search results more than 50% of the time. In fact, it seemed to show up in commercial queries (7.33% of all URLs) more often than it did for the informational queries (6.29% in the Perficient Digital data from the week of August 24th).
The bulk of the Wikipedia traffic loss appears to be driven by rankings drops, as Wikipedia’s overall representation in the SERPs changes only slightly, but the site did lose many of its #1 and #2 ranking positions.
This prevalence of Wikipedia exceeds the level at which Google shows its own properties, especially in commercial queries where Google properties represent only about 2.3% of the URLs we examined. Read on for more details!

Brief Background

In August of this year, Moz published its latest Ranking Factors Study. I wrote a preview of this on the Moz blog, comparing data from 165,177 search results from the queries Moz monitors with 182,340 search results from queries we monitor at Perficient Digital (Perficient Digital). These results represent the top 10 listings for more than 34,000 total queries.
The Moz data focuses on commercial queries, and our data focuses on informational queries. The article I wrote on Moz focused on the differences in mobile friendliness and HTTPS adoption. Today, I am publishing data on the prevalence of Wikipedia and Google properties in the results.
We have 3 different snapshots for the Perficient Digital data over the following weeks:

  1. The Week of April 17th (Just Before Mobilegeddon)
  2. The Week of May 18th (Just After Mobilegeddon)
  3. The Week of August 24th (After the Reported Wikipedia Traffic Loss)

We have one SERP snapshot from Moz, which is focused on commercial queries. The Perficient Digital snapshots are focused on informational queries.

Detailed Wikipedia Stats

I took all of the 347,517 URLs and ran them through a program to see how many of them were from Wikipedia. Here are the raw results from that process:
Percent of URLs From Wikipedia
Rankings drops or not, to me, that was a stunning level of penetration by Wikipedia. Interestingly enough, the results from commercial queries (from the Moz data) showed even more Wikipedia URLs than the results from informational queries (from the Perficient Digital data).
[Tweet “Surprisingly, Wikipedia may show up in Google more for commercial than informational queries.”]
What Percentage of Total SERPS Does Wikipedia Show Up For?
I also took a look at whether or not Wikipedia was showing up more than once per query. The first cut at that data was to measure what percentage of queries showed at least one Wikipedia page in the top 10:
Percent Queries With Wikipedia in the Top 10
It’s fascinating to see that the commercial (Moz) queries rank Wikipedia pages more often than the informational ones do. As for the reported drop in Wikipedia presence, you can’t really see much evidence of it through this method of viewing the data.
How High Do the Wikipedia Pages Rank?
To dig a little deeper, I also took a cut at the percentages of the Wikipedia results by ranking positions. For each SERP with a Wikipedia page, here are the percentages of those that rank in the top 10 spots:

This view shows us the reason for the Wikipedia traffic loss: an overall drop in rankings for Wikipedia pages. To spot this, look at the three top lines in the chart for positions 1 and 2, and you will see that there is a gap between the April Perficient Digital data and the August Perficient Digital data.
[Tweet “Study shows Wikipedia’s loss of Google traffic caused by ranking drops.”]

Detailed Google Property Stats

In our examination of all this data, one other company appeared to perform amazingly well, too … Google. However, it was not quite as strong in SERP dominance as Wikipedia. Here it is summarized:
Google Properties in the SERPs
Which Google Properties Show Up?
In this view, I dug into the details of which Google properties were ranking. This also varied significantly between the commercial query data and the informational query URLs. Here is what the data shows:
Here is the comparable data for the informational query set (using the week of August 24th data from Perficient Digital):
Google Properties in the SERPs
The largest difference by far (percentage-wise) is with the Google Books URLs, which were nearly absent in the commercial queries (three instances found), but show up in 1.54% of informational queries (2,846 instances). YouTube penetration in the results also varies pretty significantly with 1.42% of the commercial results (2,351 instances) to 3.95% of the informational results (7,294 instances).
[Tweet “Despite recent ranking losses, Wikipedia still shows in more Google results than Google properties do.”]

Summary

The Google SERPs are ever-changing, as Google constantly tweaks its algorithm. You can see that quite clearly from the above data. That said, we can conclude the following:

  1. Wikipedia did slide a bit in the rankings
  2. Wikipedia’s presence in commercial queries is actually higher than it is in informational queries
  3. Wikipedia still shows up significantly more often than Google properties in the results
  4. Google properties show up much more often for informational queries than commercial ones

Four Minute Video Summary of This Study

Slide Deck Summary of This Study’s findings

Thoughts on “Google Still Loves Wikipedia (More Than Its Own Properties)”

  1. If Google could bypass everyone, they would. Do no evil has become, Do as much evil as you can get away with.

  2. You may be interested in my blog post on the co-dependent relationship between Google and Wikipedia. The benefit to Wikipedia from having a high Google rankings is obvious: traffic, but there is also a benefit to Google — it gets a high-quality reference source at the top of its results. Google’s users expect to Google a topic and then go to the Wikipedia page and read the first paragraph. This is what the Knowledge graph attempt to replicate.
    An interesting thought experiment is to imagine Google without Wikipedia as the top link. The lack of a Wikipedia link would reduce Google’s utility, and therefore its value.
    I believe that Google’s ranking algorithms have grown to favor Wikipedia as the “gold standard” result. This ties Google’s results to Wikipedia-like content, which is dangerous for Google, especially as people’s confidence in Wikipedia declines.
    http://newslines.org/blog/google-and-wikipedia-best-friends-forever/

  3. Mark, if you believe Google is intentionally “co-dependent” on Wikipedia ranking high, then how do you account for the drop in Wikipedia from #1 result for a significant query set, which this study confirms? That seems to contradict your theory. Of course, it is possible that the Wikipedia ranking losses were an unintended consequence of some complex algorithm change.

  4. Google is trying to wean itself off of using Wikipedia as its top result using Knowledge Graph and other techniques. However, until there is a better alternative to Wikipedia it will continue to use it to bolster its search results.

  5. @ Eric Enge
    In the event that Google could sidestep everybody, they would. Do no fiendishness has ended up, Do as much abhorrent as you can escape with.
    regards
    rakhi

  6. very interesting topic!
    i by myself was thinking about this issue and i saw drop in wikipedia links on SERP but i don’t have any data to conclude this rank reduction!

  7. no comparison b/w Google & Wikipedia as both of them are good & has their own best features… We need both of them to gather information..!!

  8. @ Eric Enge
    Great work”” “Google Still Loves Wikipedia” In the occasion that Google could avoid everyone, they would. Do no evil has wound up, Do as much detestable as you can escape with.

  9. I do agree. Though Google is very popular ( which I used everyday, Googling is the term), I still use Wikipedia when it comes in searching for other topics. Wikiping would be the right term.
    They go hand in hand.

  10. I think that’s also important: wikipedia has been around for ages now, and no matter what you say, with age comes a solid domain profile and metrics.
    It’s hard to outrank those.
    Good article, thanks for sharing.

  11. If anyone is looking for a perfect on-page SEO case-study, just take a look at Wikipedia. The internal linking structure is a truly beautiful thing.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Eric Enge

Eric Enge is part of the Digital Marketing practice at Perficient. He designs studies and produces industry-related research to help prove, debunk, or evolve assumptions about digital marketing practices and their value. Eric is a writer, blogger, researcher, teacher, and keynote speaker and panelist at major industry conferences. Partnering with several other experts, Eric served as the lead author of The Art of SEO.

More from this Author

Follow Us