Tobias runs a six-person customer-success team at a mid-size software company. Last quarter he had four direct reports miss their renewal targets. He knew one of them, Renata, was struggling. He just didn't know with what. He had her numbers. He had a vague sense she was "not quite hitting it." What he lacked was a structured conversation ready to go when she sat down across from him on a Tuesday afternoon. He winged it. She left confused, and her next quarter was worse.

What This Is Really About

Performance and coaching support is the chapter most managers think they already know. After all, you give feedback. You run one-on-ones. You write reviews. But knowing the moves and executing them well under pressure are different things. This chapter teaches you how AI can sharpen your preparation so that, when the conversation starts, you are not scrambling. You are present.

The four areas below form a progression: get the data ready, build a development plan, understand team health, then give feedback that sticks.

Why This Matters

Helping your team grow is one of the most important things you do. Feedback, coaching, skill development, creating opportunities: this is where a manager's real impact happens. It is also where the most gets left undone, because all of it is time-intensive and mentally demanding.

Most managers know they should give more feedback. They know they should coach more. They know they should help people identify growth opportunities. But those activities take time and thought, so they get deferred. Then the review arrives and the feedback is vaguer than it should be, because the manager never found the time to organize their own thinking. That is exactly what happened to Tobias with Renata.

AI helps here, but not by replacing you as a coach. That is neither possible nor the goal. It helps by absorbing some of the administrative and analytical work so you have more mental energy left for the actual coaching relationship. If you can automate the organizing, the pattern recognition, and the drafting, you end up with more time and attention for the human part: the conversation and the relationship behind it.

Be clear about the goal. You are not trying to make coaching more efficient in a way that makes it less personal. You are trying to remove the administrative friction that stops coaching from happening at all. At this level of the certification you are expected to go beyond using AI on your own tasks and apply it directly to how you lead and develop the people around you.

Preparing Performance Conversations

Most managers go into performance conversations with a mental snapshot - a few recent events that come to mind easily. That's recency bias, and it tends to hurt the people who are quietly solid and reward the people who are loudly visible.

AI can fix that. Before Tobias meets with Renata, he pastes three months of his notes, her self-assessment, and her renewal numbers into a chat tool and asks: "What patterns do you see in how this person is performing? What's missing from this picture?" The output is not the final word. It is a prompt. It surfaces that Renata's close rate on accounts under 20 seats is fine, but her rate on enterprise deals dropped 30 percent after one large logo churned in February. Tobias had not connected those dots.

Now he has a real topic. Not "you missed target" - but "let's talk about what changed after February, and how you're feeling about the enterprise segment."

What to prepare: a two-sentence goal for the conversation, two or three specific observations with dates and numbers, and one open question you genuinely do not know the answer to. AI helps you build the first two. The third one is yours.

The Feedback Redesign Workflow

The current state for many managers looks like this. You have a sense of how someone is performing. You try to write up feedback. You are not sure what to say, so it sits in draft form, and eventually you give verbal feedback that is probably worse than the written version would have been. Here is the redesigned version of that workflow, the one Tobias now runs before every review cycle.

1. Capture observations. Over a quarter or six months, note specific examples of what you actually saw. What did this person do well? Where did they struggle? Keep it factual. "In the Q2 planning meeting, they raised three strategic considerations that others missed" beats "they are strategic." Specificity matters because it makes the feedback concrete and fair, and because vague impressions are exactly what you are trying to escape.
2. Use AI to structure. Bring the raw observations to the tool: "Here is what I observed from this person over the last six months. Help me identify themes. What are they excelling at? Where are they struggling? What patterns do you see?" It will organize your observations into themes you had half-noticed but never articulated, and it will ask questions you had not thought to ask.
3. Add context and judgment. Read what came back and test it against what you actually think. Does it match? What is missing? What is simply not accurate? This is where your coaching judgment enters. The tool organized the material. You are the one who understands the person and the situation.
4. Draft the feedback. "Here is how I see this person's performance. Here is what is working well. Here is where they need to grow. Help me write feedback that is honest, specific, and developmental." The tool drafts. You review, adjust it into your voice, and add the nuance that only you have.
5. Have the conversation. You now sit down with real, thought-through feedback instead of vague impressions, and the conversation goes deeper because of it.

The payoff is feedback that is more specific, more fair, and more useful. The person walks out with a clearer understanding of how they are actually perceived, which is a rarer gift than most managers realize. And you got there through genuine thinking rather than excessive hours.

Coaching and Development Planning

Development plans often fail because they are generic. "Improve communication" is not a plan. "By June 30, lead the next two client-facing calls and get written feedback from me afterward" is a plan.

AI is useful here as a sparring partner. You describe a team member - their role, their current gap, what kind of growth you want to encourage - and ask the tool to suggest two or three development approaches. Not to copy them, but to react to them. You might look at the suggestions and think: option two is too classroom-heavy, but option three is exactly what I was missing.

The plan you end up with should sound like it was written for one specific person by one specific manager - because it was. AI gets you to a starting draft faster. You make it real.

One practical format: a single-page document with four rows - current role, target skill, two or three concrete actions, and a check-in date. Tobias builds one with Renata in about 20 minutes. He uses AI to draft it between their conversation and his follow-up email. Then he reads it, cuts the jargon, and adds one line about something she said that he wants to acknowledge. She notices that line. That is what she remembers.

The Coaching Plan Redesign

Most managers get as far as noticing that someone needs to develop in an area, mention it vaguely, and never build an actual plan. Growth stays aspirational instead of becoming actionable. The redesign fixes that with five steps.

1. Identify the growth area precisely. "This person is talented but their communication in meetings could be stronger." "They are ready to move toward leadership but need to demonstrate greater accountability." Name it in a sentence you would be willing to say out loud to them.
2. Use AI to structure the options. "Here is the growth area. What are concrete ways they could develop here? What would progression look like? What skills would they need at each stage?" You will get frameworks and paths you would not have generated alone, which is the point.
3. Create the plan together. Sit down with the person and the options. Agree on what they will work on, on how you will support them, and on how progress gets measured. Co-creation is what builds their ownership of the plan; a plan handed down is a plan they will politely ignore.
4. Build an accountability structure. Use AI to force specificity. What exactly will they do? By when? How will it be tracked? When do you check in? Vague plans do not get executed, and a plan without a tracking owner is a wish.
5. Coach along the way. This step is irreplaceably human. Regular check-ins. Adjusting the plan when reality moves. Celebrating progress. Having the hard conversation when growth stalls.

When Tobias ran this with Renata, the change was that her development stopped being a topic and became a schedule. Growth plans become systematic rather than aspirational, people know what they are working toward, accountability is clearer, and more of your team actually develops because there is structure behind the intention.

Redesigning Opportunity Identification

Most managers have a rough sense of who is ready for more responsibility, but very few think systematically about which specific opportunity would most benefit which person. The result is predictable: opportunities go to whoever asks loudest or whoever happened to be on the manager's mind that week.

1. Know your team. You already carry their strengths, interests, growth areas, and readiness levels in your head. Write them down. Knowledge that stays intuited cannot be matched against anything.
2. Know the opportunities. What projects are coming? What gaps exist in the team? What skills does the organization need built?
3. Use AI to match. "Here is my team and what each person needs to develop. Here are the opportunities coming next quarter. Who should take what?" Some pairings will be obvious. Some will be ones you would never have considered, and those are the ones worth sitting with.
4. Have the conversations. Talk to people about the opportunities that line up with where they are trying to grow. A stretch assignment paired with explicit development intent is far more effective than the same assignment handed over as extra work.
5. Set them up to succeed. Make sure they have the resources and support they need, and check in actively, especially in the early weeks of a new project or role.

Done consistently, more people get opportunities that genuinely stretch them, growth starts to feel intentional rather than random, and retention improves because people can feel that someone is investing in them.

Team Dynamics and Engagement

Managers often sense a team problem before they can name it. Someone is quieter than usual. Two people stop collaborating on something they used to handle together. Energy in team meetings shifts.

AI can help you surface what you're noticing rather than just sitting on the feeling. Try describing what you observe: "Over the last three weeks, one team member has declined two optional syncs, another submitted work later than usual twice, and I've noticed less back-and-forth in our group chat. What questions might I ask in my next round of one-on-ones?"

What comes back is a list of conversation starters - not conclusions. The AI does not know why those things are happening. Neither do you yet. The goal is to walk into those one-on-ones with better questions than "how's everything going?"

Tobias does this before a quarterly team check-in and realizes he has three distinct conversations to have - one about workload, one about a project that feels stalled, and one that is probably personal and needs to be handled gently. Having those named in advance lets him pace the week rather than react to whatever fires up first.

Redesigning One-on-Ones

One-on-ones are where coaching actually happens. They are also where a great deal of time disappears without much getting accomplished, usually because the manager arrives unprepared and the conversation wanders.

Before the meeting. You and your team member each prepare what you want to discuss. AI helps you structure your side: what is the context on each topic, what is the real question underneath the surface topic, and what would progress actually look like. Preparation quality predicts conversation quality almost perfectly.

During the meeting. This part should be fully human. Take minimal notes, just key points and commitments. Do not let note-taking pull you out of the conversation you came to have.

After the meeting. Use AI to organize what happened. What did you discuss? What did you agree on? What follow-up is needed, and from whom? Organized follow-up is the difference between one-on-ones that produce change and one-on-ones that merely feel productive.

Structured this way, the meetings get more productive without becoming mechanical. You get more out of them, and the person feels genuinely supported because the follow-through happens.

Feedback Crafting

This is where managers most often misuse AI. They paste in a situation and ask for "constructive feedback to give my employee." What they get back is polished, professional, and nobody's voice. It sounds like a performance management template, not a manager who knows someone.

The better use: ask AI to help you organize what you already want to say. You write a rough version first - even just bullet points. Then ask: "Does this have a clear observation, a specific impact, and an ask? What am I missing?" The tool can flag if your feedback is vague or if you've framed the whole thing as criticism without a path forward.

Then you rewrite it in your own words. Every sentence should sound like something you would actually say out loud in a room with that person. If it doesn't, cut it.

Consider the difference:

• AI draft: "Your documentation practices have not consistently met team standards, which creates downstream challenges for colleagues."
• Tobias's version: "I noticed the last three client handoff notes were missing the technical setup section. Kofi had to ping you twice to get it. That's creating friction I want us to fix."

Same information. One sounds like a system. One sounds like a manager who was paying attention.

Common Mistakes in AI-Supported Coaching

Five mistakes undermine everything above. Knowing them by name makes them easier to catch in yourself.

Using AI as a substitute for knowing your team. The tool can organize whatever you give it, but you still have to know the people: their motivations, their history, the pressure they are under right now. AI-organized feedback built on shallow observation is still shallow feedback. Do not use the tool to avoid the work of actually understanding someone.

Giving people feedback you have not thought through yourself. The organizing can be assisted; the observation cannot. Never treat model output as feedback. Treat it as thinking material that you then process and validate with your own judgment before a word of it reaches the person.

Building so much structure that coaching becomes mechanical. Coaching is a relationship. Structure exists to support that relationship, not to stand in for it. If your team member feels like they are being processed through a system rather than developed by a person, something has gone wrong regardless of how good your templates are.

Assuming a development plan will work without real accountability. You write a well-designed plan and then nothing happens. Someone has to own the tracking and the follow-up, and that someone is usually you. The plan is only ever as good as the accountability structure behind it.

Disconnecting growth work from real work. The best development happens when people stretch on projects that genuinely matter to the organization. Plans that live independently of the actual work start to feel like homework. Tie growth to what the team is really doing.

A practical example of getting this right: a manager uses AI to draft performance review templates from each team member's quarterly goals and accomplishments, then adds their own observations and context on top. The drafting cuts preparation time roughly in half while the feedback stays meaningful and human, because the human part was never delegated.

Putting It Together

These four areas work as a loop. You prepare better conversations, which surface development needs, which let you track team health accurately, which make your feedback sharper and more credible. None of it is about delegating the thinking to an AI. It is about doing the thinking with better inputs and a clearer structure before you sit across from someone whose career you have real influence over.

Tobias's team had a better quarter. Not because of AI. Because Tobias stopped winging it. He knew what he wanted to say, why he was saying it, and what outcome he was hoping for before the meeting started. The tools helped him get there faster.

What AI-Supported Coaching Actually Produces

When you redesign feedback, coaching, and development this way, the returns go well past time saved.

Feedback becomes more specific, more fair, and more useful, because you organized your thinking and found the patterns instead of writing from impressions. People receive a clear picture of how they are actually perceived, which most managers fail to give simply because it is hard to articulate under time pressure.

Development plans become actionable rather than aspirational, because they are structured and backed by real accountability. More people genuinely develop, because there is something holding the intention up.

Your team feels invested in, because their growth looks deliberate rather than accidental. That shows up in engagement and in retention. People who can see that their manager has a real plan for them stay longer and perform better.

And you become a better coach. Not because a tool replaced your judgment, but because it cleared away the administrative friction that was quietly preventing your coaching from happening at all. The human parts of the job, the relationship and the conversation and the judgment, become more present once the clerical parts are handled.

Getting Started

Do not try to redesign all of feedback and coaching at once. Start with your next feedback conversation and run the structured workflow through it. Then build a real development plan with whoever on your team has been asking about growth. Then add preparation and structured follow-up to your regular one-on-ones. Notice what changes and what still feels awkward, and iterate from there. Tobias rebuilt his practice one conversation at a time, and the quarter turned before the system was finished.

Practice and Reflection

Take a few minutes with your own team in mind. Who on your team is closest to being ready for a bigger stretch? What would genuinely support their growth right now? What specific feedback would actually help them this week, not in six months? Which upcoming opportunity would develop them most effectively?

Now imagine having a structured process for thinking all of that through, consistently, for every person who reports to you. That is the opportunity in this chapter.

Related Lessons

The four lessons in this chapter each take one part of the loop and go deep on it.

• Preparing Performance Conversations is where the recency-bias problem gets solved in detail: pulling the full picture together so you arrive with observations rather than impressions.
• Coaching and Development Planning takes the coaching plan redesign further, turning a growth area into a co-created plan with named actions, dates, and an accountability structure.
• Team Dynamics and Engagement extends the work of naming what you are noticing across the whole team, so that the signals you sense become questions you can ask.
• Feedback Crafting focuses on the last mile: turning organized thinking into words that sound like you and land with the person in front of you.

Key Takeaways

• Prepare the conversation before you have it. Use AI to spot patterns in performance data so you enter one-on-ones with specific observations, not vague impressions.
• Development plans need to be specific to one person. Use AI to generate options, then react to them. The final plan should be concrete: named actions, named dates, named outcomes.
• Name what you're observing before interpreting it. When team dynamics feel off, describe the signals to a tool and generate better questions for your next round of check-ins.
• Write your feedback first, then let AI pressure-test it. Ask whether your draft has a clear observation, a specific impact, and a path forward - then rewrite in your own voice.
• AI helps you prepare; it does not replace the conversation. Every tool in this chapter is a preparation aid. The actual performance and coaching work happens between you and another person.
• Recency bias is a real risk in performance management. Systematically pulling three months of data before a review prevents you from over-weighting the last two weeks.
• Feedback that sounds like you is feedback that lands. If a direct report could not tell whether the message came from you or a template, you have more rewriting to do.