Coaching from QA results — closing the loop
QA without coaching is data, not improvement
A QA score that doesn’t lead to coaching is wasted effort. Most operations have the intent — "we coach from QA" appears in every operating model — but lack the structure that makes it happen. Evaluators score on Wednesday, the report lands on Friday, coaching happens "when there’s time," and most weeks there isn’t. The behaviour doesn’t change. The score doesn’t move. The programme produces data and not improvement.
The structure that makes coaching happen
Four elements separate operations that coach from QA from operations that intend to.
1. Within-week cadence. Coaching happens within 5 working days of the score landing. After that the contact is forgotten, the agent has moved on, and the conversation is abstract. The week is the maximum useful window.
2. One behaviour at a time. A coaching conversation tries to land one change. Multiple changes confuse the agent and dilute the focus. Pick the highest-leverage item from the evaluation and coach that. The next session covers the next item.
3. Specific evidence. “Your call handling needs to improve” is not coaching. “On the call at 11:23 yesterday, when the customer asked about X, you did Y; the better move would have been Z, because...” is. Specific evidence is what makes the conversation actionable.
4. Documented follow-up. What did the agent agree to try? When will it be reviewed? Without follow-up, the coaching conversation is a one-off; with it, it’s the start of a development cycle.
Performance coaching vs developmental coaching
The two are different conversations and often get conflated. Performance coaching addresses a specific deficiency that’s affecting outcomes — the agent isn’t doing X consistently. Developmental coaching helps a good performer grow — the agent is solid and ready to take on Y.
Performance coaching tends to be evaluator-led: here’s what I observed, here’s what needs to change. Developmental coaching tends to be agent-led: where do you want to be in 12 months, what do you want to work on, how can I help? Operations that run only performance coaching plateau their good agents; operations that run only developmental coaching leave performance issues unaddressed. Both are needed.
The simple structure that lifts the hit rate
Three parts. One behaviour — what specifically should change. One example — a recent specific call or contact illustrating the gap. One action — what the agent will try, by when, and how it will be reviewed. The structure takes the coaching conversation from 30 minutes of analysis to 15 minutes of practical agreement. It also makes the follow-up conversation possible: did the action happen, did it work, what next.
Common pitfalls
Coaching that’s really telling. The manager explains what should change; the agent agrees; nothing happens. Effective coaching uses questions to draw the answer out of the agent. See coaching skills for contact centre managers.
Bundling everything into one conversation. The agent left the room knowing seven things need to change. They’ll change none of them. Pick one.
No follow-up. The coaching conversation happens, the agent commits to action, nobody checks. The next coaching conversation starts from scratch. The discipline of follow-up is what separates coaching cultures from operations that "coach."
Operating model considerations
Team leaders typically own coaching from QA results. The constraint is time — team leaders are stretched, and coaching slots get squeezed out by reactive demands. Operations that take coaching seriously protect the time explicitly, often with calendar blocks that operations leadership defends.
The other operating model question is where the QA lead and evaluators fit. The cleanest model has them as advisors — surfacing the item to coach, providing the evidence, possibly running the calibration sessions for team leaders on coaching itself, but not owning the coaching conversation. The team leader owns the relationship and the change.
Conclusion
Coaching is the loop that closes the QA programme. Without it, QA produces data. With it, QA produces improvement. The discipline is in the structure — within-week cadence, one behaviour at a time, specific evidence, documented follow-up — not in the scoring methodology or the platform. Operations that take coaching seriously make QA worth doing; operations that don’t produce scores nobody acts on.
Pair this with designing a meaningful QA programme, coaching skills for contact centre managers, and calibration done well..