Volume Measures and Their Failure Mode
Counting things done is the default measure. It works, within limits, and the limits are predictable.
Tickets closed, units produced, cases handled, calls answered. Volume is the most common productivity measure because it is the easiest to collect.
The measurement warning in “Volume Measures and Their Failure Mode” matters whenever software records work patterns. Organisations researching time tracking with screenshots can use the software website for time and project context, while outcomes, quality checks and direct feedback remain necessary to explain what the metric cannot show.
Where it works
Comparable units, similar effort, high volume.
For an independent perspective related to “Volume Measures and Their Failure Mode”, consult the NIST Privacy Framework; it offers a useful external check on definitions, governance and the assumptions built into a proposed measure.
Processing, manufacturing, handling, fulfilment.
Here volume is a sound measure and the discipline around it is mature.
The failure mode
Volume rises by taking easier items first, by splitting work into smaller pieces, by closing things that are not resolved, and by declining the difficult.
None of these requires anybody to act dishonestly.
Each is a reasonable local response to being measured on count, which is the whole of the gaming problem.
The cherry-picking pattern
Where people choose their next item, they will choose the quick one.
Which raises the count and lengthens the queue for everything hard.
Visible as: rising volume, rising age of the oldest open item, and that second figure is the detector.
The splitting pattern
One job recorded as three.
Common where the unit is defined by the team, which the unit-problem note covers.
Visible as: rising volume, falling average size, which is worth tracking alongside.
The premature closure pattern
Items closed before they are finished, reopened later, closed again.
Counted twice, resolved once, and the customer experienced it twice.
Visible as: rising volume with rising reopen rate, which is the clearest single counterweight available.
The counterweights that work
Reopen or rework rate.
Age of the oldest open item.
Average size or complexity, however crudely captured.
And customer-side measures where they exist.
Any one of these makes volume much harder to inflate, and all four together make it close to honest.
Reporting volume properly
Never alone.
With its counterweight on the same page, same period, same chart where possible.
And with a note of what counts as a unit, because that definition does more work than the number.
When volume is going up for a good reason
Demand rose. Process improved. A bottleneck was removed.
These look identical to gaming in the headline figure and completely different when the counterweight is beside it.
Which is the practical argument, independent of anybody's motives.
What to check
Is your volume measure paired with anything?
What is the age of your oldest open item, and is it rising?
Has average item size fallen since measurement started?
And can people choose which item to take next?