Does anyone have a benchmark on case deflection due to your Knowledge Management or self-service capabilities?

Isaiah1 · ‎01-10-2019

Hi everyone, Isaiah here and I am new to ServiceNow and this community. I am after some KPIs and benchmarks.

In particular:

1) How are you measuring case deflection i.e. your employees don't need to raise a case thanks to your knowledge articles?

2) What kind of % deflection are you seeing or any industry benchmark you might have?

Thanks in advance!

Isaiah1 · ‎02-15-2021

Hi @Paulsylo ,

We created a widget linked to a table that does the following:

A = How many unique users read any knowledge articles in a 24 hour day
In A, how many of these unique users raised a case = B
In A, how many of these unique users raised a live chat = C
Deflection = (A - B - C)/A

The assumption is that if someone is reading a knowledge article they are not raising a case so that's a deflection.

View solution in original post

Gaurav Bajaj · ‎01-10-2019

Well, there is no benchmark as such which can give you details how many incidents got deflected because of knowledge article. Although if the user is satisfied with the article, they can mark it helpful there itself and then one can report on kb_feedback table to get the revelant data).

But there is nothing explicitly defined for it.

Thanks

Gaurav

Isaiah1 · ‎01-14-2019

Thanks Gaurav.

eric_hemmer · ‎01-11-2019

Hello - Our HR Customers typically report 50% to 80% case deflection when they make quality knowledge articles available to employees through the Employee Service Center portal. There are good webinars and case studies from Magellan Health, Christus Health, and Hitachi Ventura on the ServiceNow corporate website that you can access. Magellan Health reports 80% case deflection and 100% of cases come through their portal. Like many customers, they discontinued email as a channel for raising cases.

In terms of measuring case deflection, the easiest way is if you know how many cases you were processing before releasing the portal and knowledge and then comparing. If you were not counting cases prior to releasing the portal and knowledge, it gets trickier. There are a lot of opinions for how to measure case deflection. None of them are perfect, but here is the best in my opinion. It will require some scripting in a scheduled job. Compare employees knowledge article views to cases they raise. If an employee views an article and subsequently does not open a case within 24 hours of viewing that article, then it counts as a case deflected because the employee answered his/her question via self-service.

If anyone would like to submit a script for the scheduled job I've described, I think that would be widely helpful.

Isaiah1 · ‎01-14-2019

Thanks Eric, this is interesting as an approach. Do you know if other clients have tried to script this?