Google/Search and Wikiversity

This learning project explores the topic of search engine optimization by analyzing the results from Google Search Console.

User:Mu301 has signed up for a Google Webmaster account and this page will be used to document what he has learned about how Google searches and indexes the en.wikiversity.org site and how results from these searches bring people to Wikiversity. You are encouraged to join this learning project by discussing and analyzing the results.

Data and reports
Data for Google searches of Wikiversity was downloaded on the following dates. Each snapshot includes four weeks of results prior to the date. The index status file contains one year of data.
 * Sep. 1, 2009
 * Dec. 31, 2015
 * Jan. 8, 2018

Reports that are part of this project and the associated data dumps are in subpages. The results of these reports are summarized on this page.

Search indexing
Google periodically uses a "web crawler" to scan the pages on our site, rate, and index them. It then uses this information to produce results for a search term that may or may not include Wikiversity pages. The "index status" indicates how many unique URLs are indexed by Google. Duplicates (such as redirects) are not counted.

What is the index status of wikiveristy?
Google has reduced the number of our pages that it is indexing to ½ of what it was just one year ago. If we extrapolate this trend forward it looks like we will hit zero sometime around the middle of the fourth quarter 2018.

According to Special:Statistics we have 25,399 content pages. (Not including talk pages, redirects, etc.) On Dec. 31, 2017 there were 3,921 URLs on our site that Google indexed. That is only 15% of the total. The scope of Google's indexing is unclear. Our robots.txt file only excludes a few select Wikiversity: namespace pages and all of Draft: and User: namespaces.



The data plotted in the graph above is available at Google/Search and Wikiversity/2017/Index status.

Search queries
When someone loads a page of search results that includes a Wikiversity page this is called an "impression" and if they visit a page listed in the results it is called a "click." The "Click Through Rate" (or CTR) is the number of clicks divided by impressions. The "position" is how far down the list of results our page appears. There are typically 10 positions per page of search results.

Which search queries return Wikiversity pages?

 * Dec. 31, 2015: In the previous month Wikiversity has made between 75 to 175 impressions with 0 to 10 clicks through to our site.
 * Jan. 8, 2018: In the previous month Wikiversity has made between 4 to 45 impressions with 0 to 1 clicks through to our site.

There has been a significant drop in activity from Google search to Wikiversity resources in the past two years.

Below are the "top ten" search terms that return Wikiversity in the results. In 2015 1000 Songs/TALLIS' CANON (Thomas Tallis) had the most impressions and brought the most clicks to our site, namely 2. In 2017 MCAT Study Academy had the most impressions, but brought no activity to our site. None of the search queries resulted in more than 1 click. UFO and subpages have brought us 18 "alien sightings" but has since been prod'ed. The deleted Parapsychology pages had 7 impressions, but brought no search traffic to our site.

What is remarkable below is not what is included, but what is missing. Some of our best resources don't even appear in the top few hundred.

The tables in this section are sorted by impressions.



Which search results are people clicking on?
The tables in this section are sorted by click through rate and omit results that have zeros or negligible results. For 2009 clicks and impressions are not available.

Here we also see a clear trend that there is far less traffic from Google search reaching our site recently. The decline in clicks and the position of our pages in search results has been substantial.



Which pages at Wikiversity are most linked to?
There is a clear downward trend from 2009 to 2015. I'm not sure how to interpret the 2017 results. Has Google stopped counting external links to our site? It is also notable that Wikipedia is not included in any of these results. There should be a decent number of incoming links from there. Note: the source domain for the one recent entry is wikiversity.org itself.

