A coaching 1-on-1 template is a repeatable agenda for the regular conversation between a team lead and an agent, structured so that it changes behavior rather than just reviewing numbers. The common failure is a 1-on-1 that drifts into a status update, a metrics recap, or a vague pep talk, none of which an agent can act on. A good template does three things every time: it grounds the conversation in specific evidence from real interactions, it focuses on one behavior rather than a list, and it ends with a concrete, agreed action the agent owns. The structure below is a starting point you can adapt to your team and cadence.
In short
- Most 1-on-1s fail as coaching because they review numbers instead of changing a behavior.
- A coaching session should be grounded in specific evidence from real conversations, not in an aggregate score.
- Focus on one behavior per session; a list of five things to fix changes none of them.
- End every session with a single concrete action the agent agrees to and owns.
- Balance the conversation: reinforce what is working, not only what is not, or agents brace for every 1-on-1.
- The template is only as good as the evidence behind it, which is why it works best when scores trace back to real conversations the agent can see.
Why most 1-on-1s do not coach
Ask a team lead what happens in their 1-on-1s and you often get a description of a status meeting: here are your numbers, here is what is coming up, any blockers, see you next week. That is a useful conversation, but it is not coaching, and calling it coaching is why so many agents feel they are measured constantly and developed rarely.
Coaching changes a specific behavior. That requires a different structure from a status update: it has to start from evidence of what the agent actually did, isolate one thing worth changing, and leave the agent with a concrete action rather than a general exhortation to do better. Without that structure, even a well-meaning manager defaults to reviewing the score, and reviewing a score changes nothing, because a score is a summary and summaries are not coachable. The handoff from QA scores to a coaching conversation is covered in turning QA data into coaching; this is the structure for the conversation itself.
The coaching 1-on-1 agenda
A reusable structure. Adapt the timing to your cadence, but keep the order, because it moves from evidence to action rather than the reverse.
| Segment | Purpose | What it sounds like |
|---|---|---|
| Open: how are things | Check in as a person before a performer | A genuine question, not a formality |
| Reinforce: what went well | Anchor on a real strength with a specific example | Here is a conversation you handled really well, and exactly what you did |
| Focus: one behavior | Introduce the single thing to work on, with evidence | Let us look at this conversation together, here is the moment I want to talk about |
| Explore: the agent’s view | Understand why it happened before correcting it | What was going on here, what would you do differently |
| Agree: one action | Leave with a concrete, owned next step | So this week you will do X, and we will look at how it went |
| Close: support | Confirm what the agent needs from you | What would help you make that change |
The rules that make the template work
The agenda is only half of it. Three rules determine whether the session actually lands.
Ground it in evidence, not aggregates
Do not open a conversation you cannot show. Bring the specific interaction, and look at it together. An agent can argue with a score of 78; they cannot argue with a transcript in front of both of you, and they do not feel ambushed by a number whose origin they cannot see. This is why coaching works best when every score traces back to the conversation that produced it.
One behavior, not a list
The temptation is to cover everything since last time. Resist it. A session that raises five issues sends the agent away overwhelmed and changes none of them. Pick the one behavior with the most leverage and let the rest wait.
The agent talks more than you do
A session where the manager talks for most of it is a lecture. Ask what the agent saw and why they made the call they did, before offering the correction. People act on conclusions they helped reach, which is the principle behind any real coaching framework.
Finding the right thing to coach
The template assumes you already know which behavior to focus on, and that is the part managers most often get wrong, because they coach from the conversations they happened to see. If your view of an agent is a handful of sampled interactions, you will coach whatever was in that handful, which may not be the behavior most worth changing.
Coaching from the full picture is different. When an agent’s conversations are scored across all of them rather than a sample, a pattern becomes visible: the specific criterion this agent consistently misses, the situation that repeatedly trips them up, the strength worth reinforcing. That is what tells you which single behavior to bring to the 1-on-1, and it comes with the exact conversations attached as evidence. Kaizo produces coaching recommendations grounded in that full-coverage scoring, so the manager walks into the session already knowing the one thing worth focusing on and holding the evidence to discuss it. Proof it works at scale: at EverHelp, teams spent 75% less time preparing coaching sessions with Kaizo, because the preparation, finding the behavior and the evidence, was already done.
Frequently asked questions
What should a coaching 1-on-1 include?
A repeatable structure: a genuine check-in, reinforcement of a specific strength with an example, a focus on one behavior grounded in real evidence, exploration of the agent’s own view before correcting, a single agreed action the agent owns, and a close confirming what support they need. The order matters because it moves from evidence to action rather than opening with a score.
Why do most 1-on-1s fail as coaching?
Because they review numbers instead of changing a behavior. A typical 1-on-1 recaps metrics, previews the week, and ends, which is a status update, not coaching. Coaching requires grounding the conversation in specific evidence, isolating one behavior to change, and leaving the agent with a concrete action. Without that structure, managers default to reviewing a score, which changes nothing.
How many things should you coach in one session?
One. Raising a list of issues sends the agent away overwhelmed and changes none of them. Pick the single behavior with the most leverage, coach that with real evidence, agree an action, and let the rest wait for future sessions. Focus is what turns a 1-on-1 into behavior change rather than a review.
How do you know what to coach an agent on?
From the full picture of their conversations, not the few you happened to see. Coaching from a small sample means coaching whatever was in that sample, which may not be the highest-leverage behavior. When conversations are scored across all of them, the consistent pattern becomes clear, along with the exact conversations to use as evidence, so you bring the right single behavior to the session.
Related terms
Walk into every 1-on-1 knowing the one thing to coach
Tell us how you prepare for coaching sessions today. We will show you how full-coverage scoring surfaces the single behavior worth focusing on for each agent, with the exact conversations attached as evidence, so your 1-on-1s change behavior instead of reviewing a number. Teams like EverHelp cut coaching prep time by 75% this way.
