Two Channel Averages Look Simple. Why Is the Combined Average Wrong?
Averaging channel averages can mislead when case volumes differ. Use total handling time and the matching completed-case count to check the combined result.

Email handling averages ten minutes. Phone handling averages five. The combined average must be seven and a half minutes, surely?
Only if the two averages represent equal numbers of cases. In a small support operation, they often do not. A neat calculation can give a misleading answer because it treats a small group and a large group as equally important.
Ask what sits beneath each average
Consider a fictional service desk. One channel completes ten cases in one hundred minutes; another completes ninety in four hundred and fifty minutes. Their averages are ten and five minutes.
Across both channels, the recorded handling time is five hundred and fifty minutes for one hundred completed cases. Under that definition, the combined average is five and a half minutes. The larger group contributes more cases to the result.
This example is deliberately simple. Use your own agreed definition of handling time and a matching case count. If one channel includes follow-up work and another excludes it, correct that mismatch before combining anything.
Keep the numerator and denominator together
Record what the total time includes and which completed cases it belongs to. Do not divide one period’s time by another period’s case count. Do not combine working time with elapsed waiting time and leave the label unchanged.
When reviewing a report, inspect the underlying totals as well as the displayed average. A number that cannot be connected to its components is difficult to explain when somebody asks why it changed.
If there are no eligible completed cases, decide how the report should communicate that situation. A blank, a note and a genuine zero can mean different things. Choose deliberately rather than allowing an empty period to look like exceptionally fast service.
Test a deliberately uneven example
Use a small set of fictional channel totals with noticeably different volumes. Calculate the combined result independently, then compare the report with that control.
Repeat the check after selecting one channel or a shorter period. The expected result should follow the selected scope. If the report behaves differently, investigate the calculation and filtering before using it to discuss performance.
Keep the control example with the validation note. It helps the next maintainer understand the intended calculation without reverse-engineering a polished visual.
Do not ask the average to describe every case
A correct combined average still does not show the longest wait, the most difficult case or the distribution of work. Use it for the question it answers and keep other operational questions separate.
The useful improvement is modest: one average that can be traced to the right totals. That gives the owner a defensible starting point for a conversation, rather than a reassuring midpoint between two unrelated numbers.
Continue with a related question
Explore Make Service Numbers Useful for Your Next Business Decision, or read Your Dashboard Is Green. What Does the Number Actually Mean?.
For a written practice option, see Power BI with Matze: contents and requirements.
Review the provider’s course details. Affiliate link: Superstore may earn a commission if you buy.




Discussion 0
What did this article make you think about?