Skip to main content
ExLibris
  • Subscribe by RSS
  • Ex Libris Knowledge Center

    How to – Test the actions recorded for Primo Analytics in live Primo

    Created By: Stacey van Groll
    Created on: 2/10/2020



    Some of the descriptions for Primo Analytics Actions Usage are unfortunately quite cryptic, or missing completely from the documentation, and it can make you uncertain of what you’re measuring, which is never a good place to be with statistics!

    You can open a SalesForce case to Ex Libris but it’s much easier and quicker to start with having a look and seeing if you can figure it out yourself.

    Recorded Actions Usage data to Primo Analytics occurs when Ex Libris have correctly developed a beacon associated with that action, and it’s firing as expected when that action is initiated.

    This can be seen in the browser's developer console, upon clicking on the action.

     

    Tip!  Always be mindful of the real likelihood of skewing your own data during testing, especially with low use features.  You might not think it’s much, but those clicks really add up!  If possible, set up a test view to check such things out, which is also very handy for testing in general of Views and Scoping type related areas.  Still be aware that some data can’t be filtered out effectively by View or User Group and not affect the results, like a missing Rank in the Popular Searches Subject Area.

     

    Workflow steps (screenshots below):

    • Go to your Primo test view and navigate to the area you’d like to test
    • In Chrome, click F12 on your keyboard to open the developer console
    • Wait until your page finishes loading
    • In the Network section, click on the Clear icon
    • Do the action you’re testing, with the example I’m using being the Send To Email
    • One of the very first lines in the Network option will start with: apache_pb…
    • Click on this line
    • In the new section display which opens, click on the Headers option
    • The Request URL at the top actually has the information you want, but it’s easier to scroll down to the Query String Parameters, where it’s all parsed out
    • See various elements in the Query String Parameters which will be recorded, with some notable ones copied here
      • G: LIBRARYSTAFFB ---- User Group
      • I: 61UQ --- Institution
      • O1: blended --- Search scope
      • O21: NUI --- Primo New UI
      • OP: E-mail --- The action I just clicked
      • P: sendTo --- The group of that action

     

    Note that some actions will generate more than one apache line by design, such as Citation and then position eg Citation2.  This is as expected, but it’s also important to be aware of this when reporting, as you could build a false story from the data which actually includes duplicated counts.

     

    Tip for more testing – If you think that the data isn’t being recorded correctly even though the beacon appears to be firing, gather proof for a SalesForce case via in Analytics by live testing in UI.  Using a test view, very carefully initiate the actions you expect to be recorded, making sure you only trigger those actions as much as possible, writing down everything you click, in what order, and taking screenshots and video if possible.  The next day, after Primo Analytics has performed the nightly ETL (Extract, Transform, Load), run a report limited to your test view and the date, and see if the recorded actions match what you actually did.  PS Also remember that there may be other scenarios where the beacon is firing or not firing.

     

    Getting ready to initiate the action

    After clicking on the action

    Finding the correctly recorded action

     

    OLH and Guides (which doesn’t include this information on beacons):

     

    Additional CKC, where I’ve made attempts to figure out some of the more mysterious actions:

     

    Feedback and suggestions: