Consequences of and/or reasons for changing KB Property glide.knowman.view_age.days from 30 default to 0

Michael QCKM
Tera Guru

Re: Number of days (integer, default 30) used when summing article views (glide.knowman.view_age.days)

I know I've asked this before, but it keeps coming up and it is NOT clear to me or anyone in my group.

We need to understand the consequences of this field when creating Reports and looking at List Views.

  • If I run a Report to see the number of KB articles, never Viewed, Last updated over 1 year ago, I get a big list, but I'm told that it is wrong because it is ONLY really showing the last 30 days.
  • We try to show historical upkeep, like Drafts in queue for over 6 months, or articles in Review for over 6 months, etc.  But everytime I give a report - that looks good - nobody trusts it because they think it's actually only calculating 30 days despite what it seems.

  • If we set this to 0 to show ALL days, is there a negative impact of this?
    • Is anyone doing it this way?
  • If I leave it at 30 days, can I provide accurate historical/trend reports for periods LONGER than 30 days?
Number of days (integer, default 30) used when summing article views (glide.knowman.view_age.days) Enter a number of days to consider when calculating view count. The View count field in the Knowledge form will be updated periodically to reflect the view count based on this property. For example, if you enter 60, the view count is the total number of views in the past 60 days. Enter 0 to have ServiceNow consider all views, regardless of date.
Note: If you edit this property, run the Count Knowledge Use scheduled job to see the changes immediately. By default, the Count Knowledge Use scheduled job runs daily and updates the view counts on the knowledge articles. To view the scheduled job, navigate to System Scheduler > Scheduled Jobs > Scheduled Jobs.

New customers as of the Paris release will not have access to the View count property and Count Knowledge scheduled job. Existing customers prior to the Paris release can continue to use the property and scheduled job.

If you set the property value to 0:

  • The Count Knowledge scheduled job does not run.
  • The View count of an article contains the accumulated value of total count.
2 ACCEPTED SOLUTIONS

Ah, yes, you can see the ones that weren't viewed in the last 30 days looking at the records with "sys_view_count = 0", but those might have been viewed before that.

It's hard to report on things that didn't happen, in your case, the easiest way to find articles that have never been used is to make list of KB numbers from your kb_knowledge table, another list of unique numbers in kb_use, and diff both lists. Very manual, but I don't see any other quick solution.

I hope this helps.

Kind regards,
Lucas Vieites
P.S. If my reply helped you, please mark it as helpful or correct so other community members can benefit from this information.

View solution in original post

Tommy SN Sahlin
Kilo Sage

Hi @Michael QCKM , @Lucas Vieites and @Chris Armstron1 

 

I just stumbled over this post which has been quiet for over a year, but if I understand correctly what you are trying to achieve report-wise, I think you can do it quite easily in Report Designer.

 

  1. Create a new report based on the kb_knowledge table, set other conditions as needed
  2. Click Related List Conditions at the bottom of the Conditions section
  3. Click on the phrase "Greater than or Equal to 1", change it to "None"
  4. Click the -Table- dropdown, select "Knowledge Use->Article", set other conditions as need

Once done it should look like this:

TommySNSahlin_0-1677072552120.png


You have now told the report to return a list of all articles in the kb_knowledge table that do NOT exist in the kb_use table = have not been viewed.

 

cheers, hope that helps /Tommy

View solution in original post

13 REPLIES 13

Hi @Sweta1012 , as you can see in my previous reply, these fields capture different things; "Use Count" captures how many time a KB article is "used", meaning that it was attached to a case, incident, etc. On the other hand, the "View count" shows how many times the article was read (viewed) by users. Usually the "views" are higher than the "uses" because people will read an article, but not necessarily attach it to a task.I hope this helps.

Kind regards,
Lucas Vieites
P.S. If my reply helped you, please mark it as helpful or correct so other community members can benefit from this information.

@Lucas Vieites Thank you so much for replying. It is very helpful.

I have understood how the 'View_count' is working. The part I am not able to understand is the difference between 'Use_count' in knowledge table and 'Used' Field in kb_use table. 

 

As per my understanding both the fields gets incremented when the knowledge article is attached to any incident, task etc. However, I am finding that there is a difference in the count in these 2 fields for a particular article.

I am not understanding how both the fields are exactly getting populated and why there is a difference between them. Can you help me understand if I am missing something?

Hi there, I think you will find the answer in this article: KB0718592 - Knowledge View Count and Use Count; basically, use count is updated by a scheduled job, and views are updated depending on the number of days you set in the property glide.knowman.view_age.days, so it won't count all the views, just the ones in the amount of defined days.

Michael QCKM
Tera Guru

AHA!!!!! It took a meeting with my GENIUS Staff to recall one of our major issues with this;

  • Articles ONLY trigger into the kb_use table IF they are viewed or used.
  • Since we are attempting to find out which articles older than 30 days are NOT being VIEWED, we cannot use the kb_use table.
  • ServiceNow does not offer me a way to report on NON-Viewed articles - which to me as an Admin is almost MORE important than Viewed.
    (Does Performance Analytics do this? We do not have access to PA)

Setting the glide.knowman.view_age.days from 30 default to 0 will NOT help us at all in this case, and will actually cause different confusions.

Ah, yes, you can see the ones that weren't viewed in the last 30 days looking at the records with "sys_view_count = 0", but those might have been viewed before that.

It's hard to report on things that didn't happen, in your case, the easiest way to find articles that have never been used is to make list of KB numbers from your kb_knowledge table, another list of unique numbers in kb_use, and diff both lists. Very manual, but I don't see any other quick solution.

I hope this helps.

Kind regards,
Lucas Vieites
P.S. If my reply helped you, please mark it as helpful or correct so other community members can benefit from this information.