The evaluation KPI formula has a better meaning when is evaluated considering only the closed incidents:
KPI = n of closed incidents in SLA time / n of closed incidents
Total number of incidents resolved within SLA time divided by the total number of incidents
The evaluation KPI formula has a better meaning when is evaluated considering only the closed incidents:
KPI = n of closed incidents in SLA time / n of closed incidents
The KPI formula is fine as a beginning step, but you have to define what is an acceptable/unacceptable value for % incidents solved wihtin SLA time. Perhaps something like a value of 70-85% compliance is acceptable, <70% is unacceptable, and >85% is exceptional.
The period should be choosen so that the devider (closed incidents) contains at least n (20?) incidents.
If you want to apply percentages, it should be done on ‘big’ numbers only; there is no use in applying percentages over 2 closed incidents (I think).
Hi All,
indeed, it is very interesting how you handle cases when a team receives very few tickets per period (e.g. 2 or 10)? Do you use the same approach for measure (#Within time / #All resolved), or you apply some other rule? In case there are only 2 tickets for the whole period and one of them is delayed, this would lead to 50% achievement. Is this OK from your perspective?
Hi Sergey,
The approach must be the same.
I would apply statistical filters to this (and other) KPI values. If sample size is too small, and your t-criterion, p or similar indicators are not good enough, you would return “no value” for KPI instead of actual calculated value.
Please include formula for all the KPI’s it help to maintain consistency over all the users.