C
Chuck Adams
The situation
An average hides how consistent the work is. Two ticket types can share the same average completion time while one is predictable and the other ranges from an hour to several weeks. To answer "are we getting more consistent?" or "which work has outliers?", you need to see the distribution, not one number.
What that costs today
None of the widget types draws a distribution. Grid, Pivot Table and Card Number summarise each group into single values, and Scatter compares two measures. So teams export their PSA data to Excel and build box plots by hand, every reporting cycle.
What we would like
A box plot widget type, based on the boxplot charts in the eCharts library that MSPbots widgets are built on:
  • One box per group, such as ticket type, subtype or item: the box spans the 25th to 75th percentile, with a line at the median (ideally with the mean marked too)
  • A choice of whisker rule: to the minimum and maximum, or to 1.5 × the interquartile range with points beyond it drawn as outliers
  • Several boxes side by side within a group, for example one per month or quarter, so you can see whether the spread is shrinking over time
  • Hover shows the numbers behind each box: minimum, 25th percentile, median, 75th percentile and maximum
Who this affects
Anyone reporting on time-based service metrics, such as completion time, resolution time or first response time, who needs to know not only how fast the team is on average but how consistent it is.
Created by Frank Tian
·