Measurement
Automation rate
Also called: touchless rate, auto-decision rate
Definition
Automation rate is the share of cases completed without human intervention — a useful operating number and a dangerous headline when quoted without its quality partner.
Rate alone can be raised trivially by lowering the bar for what qualifies for automatic handling. It only means something paired with the outcome quality of the automated cases and the leakage they cause downstream, weeks or months after the rate itself was reported. The honest pair is automation rate plus post-automation error or rework rate on the same cohort, tracked together rather than reported as separate, unlinked statistics.
Segment before you celebrate
Eighty percent automation on the simplest third of volume is a different achievement from thirty percent across the whole book, even though both could be described with a single misleading top-line number. Report the mix — what share of total volume the automated cases represent, and how their complexity compares to the residual — so a reader can judge whether the headline number reflects genuine progress or simply reflects where the easy cases happened to sit.
Watch the queue it creates
Automating the easy cases concentrates difficulty in what remains, which raises average handling time and can look like a productivity decline in the human team even though the team is now doing objectively harder work than it was before. Leadership unfamiliar with this dynamic sometimes reads the rising handling time as evidence the automation program failed, when it is actually evidence the program worked exactly as intended and shifted the composition of the remaining queue.
A rising rate should trigger a quality check, not a celebration alone
Because rate is so easy to move by adjusting eligibility rules, a sudden jump in automation rate deserves the same scrutiny as a sudden jump in any metric that is cheap to game — a quick check of whether the routing rules changed, and a look at the downstream quality metrics for the newly automated cohort before the number gets reported upward as unambiguous good news in a leadership review.
In practice
- Always publish automation rate alongside rework and leakage
- Report by segment and by complexity, not as a single number
- Track handling time on the residual queue as a companion metric
- Treat a sudden rate increase as a trigger for a quality check, not a headline alone
