GROK: Admin: Using Reports & Pulling Analytics


There may be broken links in this article, the GROK staff has been notified and is working to resolve the issue.

This article will explain how to pull analytics for GROK & Tigerware.

The main sites are the following, hosted in Tableau, and will provide up-to-date statisics for 2023 and newer. The previous sites underneath this are for archival statistics ONLY.

Most stats you will be requested to pull are:

Total GROK article views, Month with the highest total views, Number of articles purged last fiscal year (June-July), and the 3 Highest viewed articles.

Total Tigerware downloads, Total Number of unique users, Number of unique users that were students, and the Top 4 Downloads on Windows.

Current Sites:

GROK: Reports Dashboard - https://tableau.lsu.edu/#/site/ITS/projects/327

Tigerware: Reports Dashboard - https://tableau.lsu.edu/#/site/ITS/projects/326

 

Archived Sites:

There is also a site on Hawkeye for GROK & Tigerware Download Statistics: https://hawkeyereports.lsu.edu/reports/report/USS/Tigerware/SoftwareDownloadByPlatform which is the only download reporting available. You can navigate to USS/GROK for the other statistics there. This requires a myLSU username/password login via popup.

However, Matomo includes much more advanced view statistics and has more advanced functions so it is generally preferred outside of download tracking.

For both sites, access must be requested from Brandon (or current GROK/ production support developer) before being able to use it.
 

Using Matomo

Log into https://piwik.uss.lsu.edu/index.php with your LSU username (excludes @lsu.edu) and unique password; it will not be your myLSU credentials. If you are unsure or have not set a password, use the Forgot Password option. Additionally, you may need to be connected to VPN if the login still fails off campus.

Custom Search

Click the Viewers tab in the left sidebar, then click Overview. On that page, there is a small button at the top with the current date range that can be edited. Note that if too many views result (such as a full year or more) in Matomo, it may show an error page. So the information will need to be split between two date ranges in that case to pull the full information.

The reason stats are so high for some articles is due to the default reporting screen also counting automated visits from bots which increase the numbers significantly. For a more accurate count, you will need to filter it such that any page views that lasted less than 3 seconds are not counted. Also, the first page to load is the front page of GROK, so it gets the brunt of these bot visits.

To filter out bot views in Matomo, go to the Viewers Overview, select All Visits at the top, then Add New Segment, then Actions in Visit, then Visit Duration (in seconds). Select Greater than, then type 3 in the field at the right. This should give accurate numbers as long as that search criteria is in effect.

Exporting

   TBD.

 

Definitions

Outlinks - the number of times all external links (such as to google.com, or moodle.org) were clicked.

Unique Outlinks - the number of visits that also clicked an external link. This stat focuses on the visitors who clicked an external link, as opposed to the number of times the link was clicked. This is why if a visitor clicks a link more than once, the subsequent numbers are not counted, as at that point the visitor has already been counted once. This is also the meaning of "unique" in general- focusing on the individual visitors rather than the raw numbers.

Note that if a visitor clicks two different links, they will be counted twice- once for each individual link.

Unique keywords - the number of individual keywords that were entered in all searches. So if one search had "LSU Moodle" and another had "LSU GROK", this would be 3 unique keywords overall: LSU, Moodle, and GROK. Note that this stat is heavily limited as the "Keyword not defined" result comes from search engines that do not disclose the search terms.

Unique downloads - the number of unique visitors who downloaded a certain file. Compares similarly to Unique Outlinks above. So if there were 3 for "Moodle.txt", that means 3 unique people all downloaded the same "Moodle.txt" file during their visit.

20300
12/7/2023 3:28:46 PM