Within our Delivery Time metric, you have the ability to calculate percentile values rather than just the mean average we default to.

The percentile will effectively provide you with an upper limit for your chosen value. See later in this article for an example of Escaped Defects Resolution SLAs.

What's the difference between Percentile Calculation and Outliers?

Both use the idea of percentiles when looking across a range of data, but the key difference is as follows:

Percentiles will give an actual value from the scale of selected values.

Outliers will give the mean average value of all tickets within the percentage range selected.

In the below table we take a sample of 10 tickets (ordered by duration) to show the relative values for p50 and p90 depending on the use of our Percentile or Outliers functions.

Ticket

Duration (days)

Percentile

Outliers

ABC-563

53.2

ABC-735

44.0

p.90 = 44.0

0 - 90 = 20.1

ABC-824

34.5

ABC-432

33.5

ABC-556

20.6

ABC-142

15.3

p.50 = 15.3

0 - 50 = 9.6

ABC-845

10.4

ABC-528

9.6

ABC-167

7.4

ABC-997

5.5

Please also note that when setting a Percentile, the Outliers section will be disabled and by default set to 0 - 100. This is to prevent applying a percentile to an already subsetted set of tickets via outliers 😕

Example Use Case - Escaped Defect Resolution SLAs

It is common practice for measures such as SLAs (service level agreements) to be based on meeting certain levels within a percentage of the time. Percentiles are ideal for providing this measurement.

Lets take the objective to "resolve 85% of P2 Escaped Defects within 10 days". This can be configured in the following way:

1) On your chosen dashboard select "Add a metric" from the menu in the top right.

2) Within the Delivery section, find Delivery Time. Click "ADD" and then "DONE".

3) Explore the newly added Delivery Time metric card, add the appropriate filters (ie. Issue Type and Priority) and click save.

4) Lastly, open the Percentiles section, select the Days (p85) option from the dropdown and click save.

In the below example, we've also selected a breakdown and the tickets (scatter chart) view:

Did this answer your question?