The symptom
A team closes eighteen tickets in one cycle. The board describes the work done well enough, but the review does not yet say what that work preserved, improved, or taught.
The signal
“Eighteen closed. What changed for the user?”
What’s going on
A ticket helps coordinate work: description, ownership, dependencies, state. It can also hold the expected result and what will let someone check it. Counting them tells you about activity, not on its own about usefulness. When the number of tickets closed becomes the main goal, how the work is cut up can matter more than the result. Two teams cutting the same work differently produce different numbers. That counter needs context before it can inform a decision.
Usefulness is not limited to a new feature. Maintenance can preserve a service, an inquiry can rule out a bad lead, an accessibility improvement can open up a use. Preparatory work may be necessary before any effect is visible. You have to be able to explain that link without inventing a result you have not got.
So separate what was done, what was observed, and what is still expected. “The fix is deployed” describes an action. “The case that was failing now passes” brings a verification. “Support requests should go down” stays an assumption until someone examines it.
This distinction also works with no ticketing tool at all. Someone organising an event can track the tasks done and check separately whether participants have the information they need. They do not need a complicated dashboard.
Check this
Take one finished piece of work and complete:
What it improved, preserved or taught: …
What lets us say so, or when we will check: …
If you do not have access to the result, ask the person concerned for feedback. An answer still unknown calls for a check, not a verdict on the effort spent. At the next review, look at whether that feedback changes what comes next.
From where you sit
- Product: tie activity to an expected effect, and mark the results still unknown.
- Operations: make the reliability preserved visible, even with no new service.
- Management: examine the context before comparing counters.
- Recruiting: ask for a result, a piece of learning, or useful maintenance.
To discuss
Which recent piece of work was useful without adding a feature, and how do we know?