How to prepare for an agent coaching session with QA evidence

• Start with the agent’s scores by rubric skill.
• Pick one or two focus areas that have enough conversations behind them.
• Bring two or three conversations that show the gap and one that shows what’s working.
• Write the goal and opening questions in advance.
• Log the session, the goal per area and the follow-up tasks.

Coaching from QA evidence keeps the conversation on specific, recent behavior the agent can change, instead of a manager’s general impression. It also makes feedback easier to give often. Gallup reports that only 21% of employees strongly agree they’ve received meaningful feedback from their manager in the past week, and that meaningful conversations can be as short as 15 to 30 minutes when held weekly ².

Gallup’s research also lists what those conversations tend to cover: current goals and priorities, collaboration and relationships, recognition or appreciation for recent work, and employee strengths and what they do well ². A QA-based brief gives you material for several of them: recent work worth recognizing, strengths and a clear goal.

Open the agent’s recent results broken down by rubric skill before you look at any single conversation. A skill-level view shows whether a miss is a pattern or a one-off, and it shows strengths you’ll want to name in the session.

unitQ Support QA gives managers a roster of each rep’s strengths and gaps by rubric skill, in Coach, the coaching area of unitQ Support QA. Scores come from the rubric your team owns, so the skills match the language your managers already use. If you’re still setting up that rubric, start with our guide to support QA scorecards.

Choose one or two focus areas per session, and only ones with enough conversations behind them to be fair to the agent. Keeping the list short leaves time to practice each one.

Check the sample size before you commit to a focus area. In unitQ Support QA, scores break down by contact reason, line of business, BPO vendor vs in-house team, region, manager and agent, all on the same scorecard, and small samples are flagged. Treat a gap on a contact reason the agent handled twice last month as a question to ask in the session.

Bring two or three conversations that show the focus area clearly and one that shows the agent doing something well. Real conversations make the feedback concrete, and the strong example gives the agent a model from their own work.

Pick conversations where the miss is unambiguous. If a miss takes a long explanation, the check definition may need work first. In unitQ Support QA, every scored miss links back to the conversation, so you can open the exact exchange behind each result.

Write a short brief: what’s working, the focus area, the evidence, two opening questions and the goal you’d like to agree on. Writing it down forces you to be specific and gives the session a shape.

A simple brief template:

• What’s working: one strength, with a conversation that shows it.
• Focus area: one rubric skill, with the pattern in a sentence.
• Evidence: two or three linked conversations.
• Opening questions: “Walk me through what you were weighing here.” “What would have helped in this moment?”
• Goal: one observable behavior to practice before the next session.
• Follow-up: what you’ll check and when.

In unitQ Support QA, managers get a coaching brief per agent, built from that agent’s conversations, with linked evidence. You still decide what to keep, what to cut and how to open the conversation.

Open with what’s working, then move to the focus area with a question. Ask the agent to walk through one conversation before you share your read, which gives them the chance to name the miss themselves.

Keep it short and regular. Gallup’s research found that meaningful conversations can be as short as 15 to 30 minutes when they happen weekly ².

Log what you covered, the goal for each area and the follow-up before you move on to the next thing. A coaching record shows whether a focus area is improving, and it lets the next manager pick up where you left off if the agent changes teams.

In Coach, unitQ Support QA also gives managers a way to log the session with a goal per area, open coaching tasks that close when the session is logged, and coaching history by manager. To see whether coaching worked, compare coached agents’ score movement with a baseline of agents who weren’t coached.

Use the same scorecard for BPO and in-house agents so feedback means the same thing wherever the agent sits. Then compare by vendor and team before you coach individuals. When a gap shows up across a whole vendor or queue, look at training and process for that team before you plan one-to-one sessions.

unitQ Support QA breaks scores down by BPO vendor vs in-house team, manager and agent, all on the same scorecard, so you can see whether a gap belongs to one agent, one manager’s team or one vendor before you plan the session.

AI agents aren’t part of one-to-one coaching, and their misses still need an owner. When a bot loops, gives wrong information or hands off badly, the fix often belongs to whoever designs its conversations.

unitQ Support QA grades human and AI agents side by side in one product, on as many rubrics and scorecards as your policies require, with one shared standard or bot-specific checks, independent of the vendor that built the bot ¹. AI-generated findings come with recommended next actions, each tagged with a suggested owner such as operations, enablement, QA or conversation design. The bot audit is configured for your workspace during setup. For more, read what AI agent QA is.

Here’s an illustrative prep for a chat agent on a billing queue.

• Roster view: strong on tone and next steps; below the team on identity verification across the last few weeks, with enough conversations to trust the pattern.
• Focus area: identity verification before account details are shared.
• Evidence: two conversations where the agent confirmed a renewal date before verification, and one where the agent handled an upset customer well and still verified first.
• Opening question: “In this one, the customer was in a hurry. What made it feel right to share the date first?”
• Goal: complete verification before confirming any account detail, even when the customer offers it.
• Follow-up: review the identity verification check on the agent’s conversations before the next session, then log the outcome.

QA evidence is the wrong starting point when the issue isn’t in the conversations. Attendance, wellbeing and conduct conversations belong with your HR process. New hires in their first weeks often need shadowing and side-by-side practice before scored feedback is useful. And agents who handle very few conversations may not have enough evidence for a fair skill-level view yet.