Supervisors and Team Leads: Running a Team in 2026

Lesson 1

Your role and your capacity

Who this course is for

Supervisors and team leads who manage people, set priorities and run one-to-one meetings in office, operations and retail teams. Each module has objectives, lessons, key terms, a worked example, a practice task, a 3-question quiz and a job aid.

Practice rule: examples use invented names and situations. Replace them with your own team's context.

Several figures come from secondary summaries, vendor guides and practitioner sources, and the text labels them. Check the originals before quoting any figure to your team. Advice on coaching cadence and feedback durations is practitioner guidance, not a proven formula, so try it with your team and adjust.

Objectives

• Explain why manager engagement and capacity are business constraints, not personal failings.
• Recognise span-of-control and workload signals that affect manager and team performance.
• Apply regular listening and workload review to protect capacity and reduce burn-out risk.

Lesson 1.1: Your role and the manager engagement picture

The research reports that manager engagement fell from 31% in 2022 to 22% in 2025, while global engagement fell to 20% in 2025, its lowest since 2020. In best-practice organisations, 79% of managers are engaged. The gap suggests that your role is shaped by context: support, practices and capacity matter, not just individual effort.

Capacity is not a personal failing. The research uses the occupational definition of burn-out as chronic workplace stress that has not been successfully managed; it is not a medical condition. It also reports that 27% of US full-time employees in Q2 2026 very often or always feel burned out. Five causes are largely within leaders' and managers' control: unfair treatment, unmanageable workload, unclear expectations, lack of manager support and unreasonable time pressure. How people experience workload matters more than hours alone.

Manager check-ins are a burn-out tool. Employees whose manager is always willing to listen to work problems are 62% less likely to be burned out. Manager wellbeing has also declined, with older and female managers reporting the steepest drops, according to a secondary summary. Your role therefore includes protecting listening time and reviewing workload, not only driving output.

Lesson 1.2: Span of control and team design

Span of control is the number of people one manager leads. The research reports that manager engagement declines with larger spans, though manager talent and training can offset this. Before blaming a manager for low engagement, check whether the team is too large for the role. Structure is part of performance.

Training can help. A secondary summary of 17 studies with 14,774 participants found that trained managers produced 8–18% higher team engagement and 21–28% lower turnover. Check the original before relying on the figures. Training improves capability, but it does not remove an unreasonable span. Capacity and capability both matter.

Use signals to spot span problems: missed check-ins, repeated cancellations, unclear priorities and the five burn-out causes. A manager with too many direct reports cannot give regular attention. The business constraint is time and attention. Adjust span or priorities before adding more work or blaming the manager.

Lesson 1.3: Capacity as a business constraint

Manager capacity is the time and attention available for coaching, workload review and decisions. It is a business constraint because it limits team engagement and retention. The research links trained managers to higher team engagement and lower turnover, so manager overload is not only a wellbeing issue. It carries operational risk.

Workload experience matters. Employees whose manager is always willing to listen to work problems are 62% less likely to be burned out. Regular listening conversations should track the five burn-out causes: unfair treatment, unmanageable workload, unclear expectations, lack of manager support and unreasonable time pressure. Do not wait for a crisis to discuss capacity.

Watch for perception gaps. One survey in the research found that 59% of managers believed engagement increased in the past year, while 80% of employees said it stagnated or declined. Check assumptions with team evidence. If capacity is tight, address workload, clarify priorities and escalate the constraint. Treat it as a business problem, not a personal failing.

Worked example

Priya manages 14 people and keeps cancelling one-to-ones because of workload. She thinks engagement is improving. Tom, one of her direct reports, says priorities are unclear and workload is unmanageable. Priya uses her own manager conversation to review span of control and capacity, not to blame herself. She agrees a temporary span reduction, clarifies priorities, and starts short check-ins focused on obstacles and wellbeing. She records the capacity constraint and escalates it. The example treats capacity as a business constraint, not a personal failing.

Key terms

• Span of control: the number of people one manager leads.
• Engagement (Gallup): the share of employees who are psychologically invested in their work.

Practice

Map your capacity and span: list each direct report, your recurring commitments, and the check-ins you cancelled or shortened in the last month. For each direct report, note one obstacle and one workload signal linked to the five burn-out causes. Then choose one capacity constraint to escalate and one change that protects regular listening time.

Job aid: capacity check

• List your direct reports.
• List your recurring commitments.
• List the check-ins you cancelled.
• Name one constraint to escalate.

Check what you learned

1. What does the research report about manager engagement?
   • It rose from 31% in 2022 to 79% in 2025.
   • It fell from 31% in 2022 to 22% in 2025.
   • It stayed at 20% in 2025.
2. What does the research say about span of control?
   • Larger spans always improve manager engagement.
   • Manager engagement declines with larger spans, though talent and training can offset this.
   • Span of control is the number of customers a manager serves.
3. Which statement about capacity and burn-out is supported by the research?
   • Burn-out is a medical condition and is mainly about hours worked.
   • How employees experience workload has a stronger influence on burn-out than the number of hours.
   • Manager listening has no effect on burn-out risk.

Lesson 2

Feedback and coaching

Objectives

• Give specific, behaviour-based feedback at least weekly, describing observed behaviour and its impact rather than labels.
• Separate developmental coaching conversations from judgement, rating and pay decisions.
• Use feedback evidence to check frequency, quality and effect, and avoid assuming feedback is always helpful.

Lesson 2.1: Weekly, specific, behaviour-based feedback

Feedback often fails on frequency and on agreement about frequency. The research reports that 47% of employees received manager feedback only a few times or less in the previous year. It also reports that 50% of managers say they give feedback weekly, against 20% of employees who say they receive it. These figures come through secondary summaries, so check primary sources before quoting them formally. Use them as a prompt to ask what your team actually receives. Weekly feedback should be a routine, not a rare event.

Make the feedback specific and behaviour-based. Describe what you observed and its impact, and avoid labels. The research marks this as vendor practitioner guidance, not a proven formula. The engagement link rests on meaningful, recent feedback, not on a script. For example, Priya might tell Tom: "In yesterday's handover you flagged two missing order details before they reached the customer, which reduced rework. Please keep doing that." That is observable and actionable. Vague praise or criticism gives the person little to repeat or change.

Keep a simple evidence note: date, situation, observed behaviour, impact and agreed next step. This helps you give weekly feedback and prepares fair review conversations. One reported figure says about 80% of employees who received meaningful feedback in the past week were fully engaged, against roughly one in four without it. This is a secondary summary; check the original. The practical test is whether your feedback is recent, specific and understood. If not, adjust the conversation before adding more meetings.

Lesson 2.2: Separate coaching from judgement and pay

Coaching aims to improve future work; judgement rates a person, often for pay or promotion. The research defines these as developmental and evaluative feedback. A secondary report of a 1965 study found that combining the judge role and the coach role in one conversation defeats both. So separate the conversations. If a pay decision is live, say so and move the developmental discussion to another time. Do not smuggle a rating into a coaching question. The purpose of each conversation should be clear at the start.

For example, Priya should not open a pay review with "How do you think you did?" and then coach. Instead, hold a decision conversation: state the decision, the evidence and the route to discuss it. Then book a separate coaching conversation about Tom's next skill. Keep records of the evidence used for decisions; use the same behaviour-based notes, not memory or labels. This keeps the purposes separate. Tom can hear the decision, and the coaching can focus on future work.

Coaching quality may matter more than session count. A 2026 study asks whether the quality of the coaching relationship or the number of sessions predicts outcomes; the research does not summarise the results, so check the paper. Do not assume more sessions are better. Ask the person what help they want, agree one next step, and keep evaluation out of that conversation. Decisions about people should have a named human decision-maker. If you manage both coaching and pay, make the boundary visible.

Lesson 2.3: Using the feedback evidence

Feedback is not reliably helpful. A meta-analysis of 131 studies and 607 effect sizes found a mean effect of d = 0.41, but more than 38% of effects were negative; in some cases feedback made performance worse. This is a secondary summary, so check the original. The practical point is variance. Feedback design and delivery matter. Do not assume that more feedback is better. Use specific behaviour and impact, and check whether the person can act on it. Then look for evidence that the feedback changed something.

Quality is a live problem. A 2026 survey of 2,947 employees and managers found that only 39% of employees agree their manager gives effective developmental feedback. This comes through a secondary summary; check the original. Ask for a quick check: "What feedback have you received from me in the last week, and what did you do differently?" Listening for the answer is part of the evidence. The research also reports a frequency gap, so compare what you think you give with what your team says it receives.

Build a feedback evidence trail. Record date, situation, observed behaviour, impact and agreed action. Use it to spot patterns and to prepare fair decisions. Keep it factual and share relevant notes with the person. Avoid labels. Review your own feedback monthly: is it weekly, specific and separated from judgement? If the answer is no, change one habit, such as adding a weekly note, before redesigning the whole system.

Worked example

Priya manages Tom in an operations team. On Tuesday she notes that Tom's handover included two missing order details; she records the date, observed behaviour, impact and agreed action to use the handover checklist for the next two weeks. At their weekly conversation she gives this feedback specifically and asks Tom what he will do. She does not mention the upcoming pay review. Later, in a separate decision meeting, Priya uses her notes, states the rating and its evidence, and names herself as decision-maker. She then books a separate coaching conversation on checking handover quality. This keeps coaching developmental and the pay decision evaluative, with recent, behaviour-based evidence.

Key terms

• Developmental feedback: feedback aimed at improving future work.
• Evaluative feedback: feedback that rates a person, usually for pay or promotion.

Practice

For the next two weeks, keep a feedback log with one entry per direct report per week. For each entry, record the date, situation, observed behaviour, impact and agreed next step. Hold a weekly feedback conversation using only the observed behaviour and impact. At the end, review the log: which entries are specific, which are labels, and which conversations mixed coaching with judgement or pay? Bring one example to your next supervision session.

Job aid: feedback card

• Date
• Situation
• Observed behaviour
• Impact
• Agreed next step
• Keep coaching and pay decisions in separate conversations.

Check what you learned

1. Which feedback-frequency gap does the research report?
   • 47% of employees received manager feedback only a few times or less in the previous year.
   • Most employees received weekly feedback.
   • Managers and employees report the same frequency.
2. Why separate coaching from judgement or pay conversations?
   • Because coaching should always be longer than pay conversations.
   • Because combining the judge role and coach role in one conversation can weaken both.
   • Because pay decisions should not use evidence.
3. What does the feedback meta-analysis summary say?
   • Feedback effects are uniformly positive.
   • The mean effect was d = 0.41, but more than 38% of effects were negative.
   • Feedback has no measurable effect.

Lesson 3

One-to-one meetings

Objectives

• Set a regular one-to-one cadence that protects time for development, obstacles and wellbeing.
• Run employee-led meetings with clear notes and agreed actions.
• Avoid common one-to-one mistakes, including status updates, unnecessary cancellation and mixing coaching with judgement.

Lesson 3.1: Set a cadence and protect the slot

Regular one-to-one meetings are linked to engagement: the research reports an older study finding that employees whose managers hold regular meetings with them are almost three times as likely to be engaged. Agree a cadence you can sustain. One summary recommends a short weekly quick connect of one to 10 minutes and a fuller check-in once or twice a month of 10 to 30 minutes; the research marks the primary source for these durations as UNVERIFIED.

Quality matters more than quantity, provided meetings happen at least monthly. The research reports a study finding that meeting quality mattered more than quantity for engagement, and that only 16% of nearly 15,000 employees described their most recent manager conversation as extremely meaningful. Use the time for development, obstacles and wellbeing, not a read-out of tasks. A recurring calendar slot and notes on actions help make the meeting useful.

Protect the slot. Practitioner guidance in the research says a one-to-one should be cancelled only for sickness or emergencies, and recommends a recurring calendar slot. If you must move it, reschedule rather than cancel. Frequent cancellation teaches the team that the meeting is optional. For managers with large spans of control, check workload and meeting load before blaming the manager; manager engagement declines with larger spans, though manager talent and training can offset this.

Lesson 3.2: Let the employee lead the agenda

Use an employee-led agenda. The employee sets the topics, with the manager adding only what must be covered. Practitioner guidance in the research recommends an employee-led agenda, a recurring calendar slot, notes on actions and no cancellation except for emergencies. The meeting should focus on development, obstacles and wellbeing, not status updates. Ask what support is needed and what the employee wants to change before offering solutions.

Let the employee do most of the talking. Practitioner guidance in the research suggests around 90% of the talking should come from the employee, but this is a vendor figure, not an empirical measure. Listening matters: employees whose manager is always willing to listen to work problems are 62% less likely to be burned out. That makes the one-to-one a burn-out tool, not just a performance ritual.

Separate coaching from judgement. Combining the judge role and coach role in one conversation can defeat both. Keep developmental feedback in the one-to-one and handle pay, rating or promotion decisions elsewhere, with a named human decision-maker. Record actions and follow up at the next meeting so the employee sees the agenda leading to change.

Lesson 3.3: Avoid the common one-to-one mistakes

Avoid treating the one-to-one as a status update. Status can be shared in writing; the meeting is for development, obstacles and wellbeing. Avoid cancelling the meeting except for sickness or emergencies. Avoid talking too much: practitioner guidance suggests the employee should do most of the talking, around 90%, though this is a vendor figure, not an empirical measure. These mistakes turn a developmental conversation into a report-out.

Avoid vague or rare feedback. The research reports a frequency gap: many employees receive manager feedback only a few times or less in the previous year, while half of managers say they give feedback weekly. Feedback is high-variance: a meta-analysis found a mean effect but more than 38% of effects were negative. Be specific, describe observed behaviour and its impact, and avoid labels.

Do not assume your view matches the team's. The research reports that most managers believed engagement increased, while most employees said it stagnated or declined. Check your assumptions with the employee, not just with your own impressions. Also check span and workload before blaming a manager for poor engagement. Keep people decisions with a named person.

Worked example

Priya manages Tom and holds a regular one-to-one in a recurring calendar slot. Tom sets the agenda, covering a development goal, an obstacle with a handover and his workload. Priya listens, asks what support he needs and notes agreed actions. When a diary clash arises, she reschedules rather than cancels; she would cancel only for sickness or an emergency. She keeps coaching separate from Tom's pay review and follows up on the actions at the next meeting.

Key terms

• Quick connect: a short, frequent check-in, often weekly.
• Employee-led agenda: a meeting where the employee sets most of the topics.

Practice

Draft a one-page one-to-one template with three sections (employee-led agenda, obstacles and support, and agreed actions) plus a rule that allows cancellation only for sickness or emergencies. Test it in your next one-to-one and note what you would change.

Job aid: one-to-one card

• The employee sets the agenda.
• Cancel only for sickness or an emergency.
• Notes and actions are shared.
• Follow up on actions next time.

Check what you learned

1. What does the research say about one-to-one quality and quantity?
   • Quality matters more than quantity, provided meetings happen at least monthly.
   • Quantity matters more than quality, provided meetings happen weekly.
   • Only daily meetings improve engagement.
2. According to practitioner guidance in the research, when should a one-to-one be cancelled?
   • Whenever work is busy.
   • Only for sickness or emergencies.
   • When the employee has no agenda.
3. Which is a common one-to-one mistake identified in the research?
   • Using the meeting for development.
   • Letting the employee lead the agenda.
   • Treating the meeting as a status update.

Lesson 4

Workload, burn-out and wellbeing

Objectives

• Describe burn-out using the WHO definition and distinguish it from a medical condition.
• Identify Gallup's five causes of burn-out that sit within managers' control, and spot them in your team.
• Apply listening and workload-tracking habits to respond when a team is overloaded.

Lesson 4.1: Burn-out, defined and bounded

The World Health Organization's ICD-11 describes burn-out as a syndrome resulting from chronic workplace stress that has not been successfully managed. It is an occupational phenomenon, not a medical condition, and the term should not be used for stress arising elsewhere in life. That distinction matters for managers: burn-out points at how work is designed, organised and led, not at a person's resilience. Gallup's Q2 2026 data for US full-time employees found that 27% feel burned out very often or always. That is a management signal, not a personal one.

Burn-out shows up in attendance and retention as well as wellbeing. A 2018 Gallup study of about 7,500 full-time US employees linked frequent burn-out to 63% more sick days and 2.6 times the likelihood of looking for a new job. Gallup also finds that how employees experience their workload influences burn-out more strongly than the number of hours they work. A team working long hours with clear priorities and fair treatment may cope better than a team working shorter hours under confusing or pressured conditions. Hours alone are a poor diagnostic.

Managers are not immune. Gallup's 2026 data shows manager wellbeing declined, with older and female managers reporting the steepest drops; this comes through a secondary summary, so check the primary source before quoting it. Treat your own capacity as a business constraint rather than a personal failing. Manager engagement declines with larger spans of control, though manager talent and training can offset this. That is an argument for reviewing team size and support, not for expecting managers to absorb unlimited load.

Lesson 4.2: The five causes within your control

Gallup identifies five causes of burn-out that are largely within the control of leaders and managers: unfair treatment, unmanageable workload, unclear expectations, lack of manager support and unreasonable time pressure. None of these is a personality trait, and none is fixed by telling people to be more resilient. They are features of how work is assigned, prioritised, resourced and discussed. That is why manager check-ins are a burn-out tool rather than a nice-to-have. The practical job is to spot which of the five is present, then change the condition rather than the person.

Some causes are easier to spot than others. Unclear expectations often look like busyness: people working hard on the wrong things, or waiting for a decision. Unfair treatment may surface as quiet disengagement rather than a complaint. Unreasonable time pressure can be hidden by a team that keeps delivering. Because employees' experience of workload matters more than hours, change the questions you ask. Instead of counting hours, ask whether the work is doable, prioritised and fairly allocated, and whether people know what good looks like.

Your span of control is part of the picture. Gallup's upskilling studies (17 studies, 14,774 participants) found trained managers produced 8–18% higher team engagement and 21–28% lower turnover; these figures come through a secondary source, so verify them at the Gallup original. Managers also tend to assume their view matches the team's. An AMA survey (October 2025, 1,300+ workers) found 59% of managers believed engagement had increased in the past year, while 80% of employees said it had stagnated or declined.

Lesson 4.3: Listening and responding to overload

Listening is the cheapest burn-out intervention available. Employees whose manager is always willing to listen to work problems are 62% less likely to be burned out. That does not mean solving every problem on the spot. It means people can raise a work problem without it being treated as a complaint or a performance issue. Regular one-to-one meetings give you the forum. Gallup's classic finding is that employees whose managers hold regular meetings with them are almost three times as likely to be engaged; this is an older study, so check the date before citing it.

Structure matters more than informality. A practitioner guide from Baylor University (2023) recommends an employee-led agenda, a recurring calendar slot, notes on actions, and cancelling only for sickness or emergencies. Gallup's coaching framework, as summarised by a secondary source, suggests a short quick connect of one to ten minutes at least weekly, plus a fuller check-in of ten to thirty minutes once or twice a month; these durations are UNVERIFIED because the primary Gallup page was not located.

Practitioner guidance suggests the employee should do most of the talking, around 90%, though this is a vendor figure rather than an empirical measure. Use the meeting to ask what is slowing the person down, what support they need, and what they would change first.

Worked example

Priya manages Tom and runs a one-to-one in a recurring calendar slot. Tom's workload has grown while priorities keep changing. In the meeting, Tom sets the agenda: his workload, an obstacle with the handover process, and a development goal. Priya listens, asks what support would help, and notes agreed actions. When a diary clash arises, she reschedules instead of cancelling. She checks the five causes, finds unclear expectations, and clarifies priorities with her own manager the same week.

Key terms

• Burn-out (WHO ICD-11): chronic workplace stress that has not been successfully managed; an occupational phenomenon, not a medical condition.
• Five causes within managers' control: unfair treatment, unmanageable workload, unclear expectations, lack of manager support, unreasonable time pressure.

Practice

Choose one team member. Write down which of Gallup's five causes you think is most present for them, and the evidence you have. Then plan one listening conversation for the next week: an employee-led agenda, one question about workload, and one action you will take within your control.

Job aid: overload check

• Which of the five causes is present?
• Change the condition, not the person.
• Listen first, then act.

Check what you learned

1. Which of these is one of Gallup's five causes of burn-out that managers can influence?
   • Long commute distance
   • Unclear expectations
   • Personal resilience alone
2. Why does the research say burn-out should be treated as a management signal?
   • Because it is a medical diagnosis managers must make
   • Because it is an occupational phenomenon that points at how work is designed and led
   • Because it applies to every stressful event in life
3. What influences burn-out more: hours worked, or how people experience their workload?
   • Hours worked
   • How people experience their workload
   • Neither

Lesson 5

Guiding AI use and handling difficult conversations

Objectives

• Apply approved-tool, disclosure and data rules when your team uses generative AI.
• Check AI output and keep decisions about people with a named human.
• Hold a conflict or performance conversation using specific, behaviour-based feedback.

Lesson 5.1: Approved tools, disclosure and data

Managers shape whether AI use feels safe and useful. Gallup's 2026 data found that employees who believe their manager supports AI use are 8.7 times more likely to strongly agree AI has transformed how much work gets done. Yet the research reports that 50% of CHROs in a Gallup roundtable were not confident in managers' ability to guide AI use; this item is UNVERIFIED and should be checked against Gallup directly. So set team rules: use only approved tools, and require disclosure when work product was created in whole or in part by generative AI, with a way to document use.

Data rules must be concrete. Do not put confidential company information into unapproved tools. Review AI output before relying on it. Do not deploy AI in customer-facing processes without review. These points come from Ethisphere's 2026 vendor-advisory guidance. The research also cites an August 2026 update to a privacy-and-employment framework saying safeguards should rise with data sensitivity, system autonomy, closeness to a decision and severity of impact; this is UNVERIFIED because the framework was not read. Use those four factors to decide which tasks need stricter checks and who signs them off.

Make disclosure a normal part of the work, not a confession. When AI contributes to a draft, email or report, say so and show that the output was checked. Tell staff what you monitor and why; this is general practitioner guidance with no dated source. Under the EU AI Act, AI literacy has applied since 2 February 2025 and Article 50 transparency rules generally since 2 August 2026 (from a secondary source); these apply only where the law reaches, so check EUR-Lex Articles 4 and 50 and your legal or data protection lead. Keep the team rules short and visible.

Lesson 5.2: Checking output and naming the human decision-maker

Approved tools are not a substitute for checking. Gartner's March 2026 HR survey found that 45% of managers say AI has improved their teams' work as much as they expected. That is a mixed picture, not a reason to accept output unexamined. Review AI output before relying on it, and do not deploy AI in customer-facing processes without review. Apply the risk-scaling factors: data sensitivity, system autonomy, closeness to a decision and severity of impact. Ask what data went in, how much the system decides, and what happens if it is wrong.

Decisions about people stay with a named human. Hiring, pay and discipline need a named human decision-maker, even when AI has drafted a summary or suggested options. Name that person in the process and record the human check. This is an accountability rule, not a ban on AI. Use AI for drafting or summarising, then have the named person review the facts, the policy and the impact before deciding. If AI contributed to the work product, disclose that and keep a way to document use. Do not let a tool become the invisible decision-maker.

Build a simple team routine. For each common task, agree the approved tool, the data rule, the output check and who signs off. For example, a team may use an approved tool to draft routine text but require a named person to check facts, tone and confidential data before anything is sent. Say when AI was used. If you monitor use, tell staff what is tracked and why. Review the routine when tools, tasks or risks change.

Lesson 5.3: Difficult conversations on conflict and performance

Feedback is often too rare and too vague. Gallup reports that 47% of employees received manager feedback only a few times or less in the previous year, and that 50% of managers say they give feedback weekly, against 20% of employees who say they receive it. About 80% of employees who received meaningful feedback in the past week were fully engaged, against roughly one in four without it. Only 16% of nearly 15,000 employees described their most recent manager conversation as extremely meaningful. Make feedback recent, specific and part of a regular one-to-one.

Separate coaching from judgement. Meyer, Kay and French's GE studies from 1965 found that combining the judge role and coach role in one conversation defeats both. Kluger and DeNisi's 1996 meta-analysis of 131 studies and 607 effect sizes found a mean effect of d = 0.41, but more than 38% of effects were negative, so feedback can make performance worse. In a conflict or performance conversation, describe the observed behaviour and its impact, avoid labels, and keep any pay or rating decision separate. The named human owns the decision.

Use structure and listen. Keep an employee-led agenda, a recurring calendar slot and notes on actions. Listening is also a burn-out tool: employees whose manager is always willing to listen to work problems are 62% less likely to be burned out. Check the five causes largely within managers' control: unfair treatment, unmanageable workload, unclear expectations, lack of manager support and unreasonable time pressure.

Worked example

Priya leads a small operations team. Tom uses an unapproved tool to draft a customer email and does not say so. Priya does not turn the conversation into a rating. She asks Tom what data went into the tool, checks whether confidential information was included, and explains the team rule: approved tools only, disclosure when AI contributes, output checked before use, and no customer-facing deployment without review. She names herself as the human decision-maker for any performance follow-up. In their one-to-one, Tom leads the agenda. Priya describes the observed behaviour and its impact, asks Tom for his view, agrees the required change, and notes actions. She separates coaching from any later evaluative decision and checks workload and support.

Key terms

• Approved tool: an AI tool your organisation has approved for specific work and data.
• Named human decision-maker: the one person accountable for decisions about people, even when AI has helped.

Practice

Write a one-page team checklist for a task that uses generative AI. Cover the approved tool, the disclosure rule, the confidential-data rule, the output check and the named human for any people decision. Then draft three opening sentences for a conflict or performance conversation with a team member who used an unapproved tool, using observed behaviour and impact rather than labels.

Job aid: before AI goes into team work

• Is the tool approved?
• Is there a disclosure rule?
• Is confidential data kept out of unapproved tools?
• Has the output been checked?
• Is a named human responsible for any decision about a person?

Check what you learned

1. A team member has used AI to draft a summary that will inform a disciplinary decision. What keeps accountability with a named human?
   • Use the draft as the final decision if the tool is approved.
   • Name a human decision-maker, use the AI only for drafting, and record the human check.
   • Ask the AI tool to confirm the decision is fair.
2. What should a manager do before relying on AI output for a customer-facing process?
   • Rely on it if the tool is on the approved list.
   • Review the output and do not deploy it without review.
   • Disclose it only if the customer asks.
3. Which feedback approach fits the research?
   • Combine coaching and a pay rating in one conversation to save time.
   • Describe observed behaviour and impact, and separate coaching from evaluative decisions.
   • Give feedback rarely and in general terms to avoid conflict.

Lesson 6

Final check

Ten questions from all five modules

Answer all ten, then check your answers. If you run this course for a team, a senior manager or trainer should go over any wrong answers. The pass mark is a team choice; 80% (8 of 10) is a common starting point to review after a pilot. If you miss one, go back to the module named in the question.

Check what you learned

1. What does the research report about manager engagement? (Module 1)
   • It rose from 31% in 2022 to 79% in 2025.
   • It fell from 31% in 2022 to 22% in 2025.
   • It stayed at 20% in 2025.
2. What does the research say about span of control? (Module 1)
   • Larger spans always improve manager engagement.
   • Manager engagement declines with larger spans, though talent and training can offset this.
   • Span of control is the number of customers a manager serves.
3. Which feedback-frequency gap does the research report? (Module 2)
   • 47% of employees received manager feedback only a few times or less in the previous year.
   • Most employees received weekly feedback.
   • Managers and employees report the same frequency.
4. Why separate coaching from judgement or pay conversations? (Module 2)
   • Because coaching should always be longer than pay conversations.
   • Because combining the judge role and coach role in one conversation can weaken both.
   • Because pay decisions should not use evidence.
5. What does the research say about one-to-one quality and quantity? (Module 3)
   • Quality matters more than quantity, provided meetings happen at least monthly.
   • Quantity matters more than quality, provided meetings happen weekly.
   • Only daily meetings improve engagement.
6. According to practitioner guidance in the research, when should a one-to-one be cancelled? (Module 3)
   • Whenever work is busy.
   • Only for sickness or emergencies.
   • When the employee has no agenda.
7. Which of these is one of Gallup's five causes of burn-out that managers can influence? (Module 4)
   • Long commute distance
   • Unclear expectations
   • Personal resilience alone
8. Why does the research say burn-out should be treated as a management signal? (Module 4)
   • Because it is a medical diagnosis managers must make
   • Because it is an occupational phenomenon that points at how work is designed and led
   • Because it applies to every stressful event in life
9. A team member has used AI to draft a summary that will inform a disciplinary decision. What keeps accountability with a named human? (Module 5)
   • Use the draft as the final decision if the tool is approved.
   • Name a human decision-maker, use the AI only for drafting, and record the human check.
   • Ask the AI tool to confirm the decision is fair.
10. What should a manager do before relying on AI output for a customer-facing process? (Module 5)
    • Rely on it if the tool is on the approved list.
    • Review the output and do not deploy it without review.
    • Disclose it only if the customer asks.

Lesson 7

Glossary and sources

Glossary

• Burn-out: WHO ICD-11: chronic workplace stress not successfully managed; an occupational phenomenon.
• Developmental feedback: feedback aimed at improving future work.
• Employee-led agenda: a one-to-one where the employee sets most topics.
• Engagement (Gallup): share of employees psychologically invested in their work.
• Evaluative feedback: feedback that rates a person, often for pay or promotion.
• Named human decision-maker: the accountable person for decisions about people.
• Quick connect: a short, frequent check-in.
• Span of control: number of people one manager leads.

Sources

Many figures come from secondary summaries, vendor guides and practitioner sources, and the lessons say so; check the original before quoting them. The EU AI Act references apply only where the law reaches.

• 1. Gallup, State of the Global Workplace 2026 (data for 2025), via HR Dive (2026); check Gallup directly.
• 2. Gallup, manager engagement and best-practice figures (2026); gallup.com.
• 3. Gallup upskilling studies (17 studies, 14,774 participants), via secondary summary; check the Gallup source.
• 4. Gallup Q2 2026 burnout data (US full-time employees); gallup.com.
• 5. Gallup 2018 burnout study (about 7,500 US full-time employees); gallup.com.
• 6. WHO, ICD-11 burn-out as an occupational phenomenon (28 May 2019).
• 7. Gallup, classic one-to-one finding (older study; check date).
• 8. Baylor University, manager best practices for one-to-one meetings (2023), practitioner guide.
• 9. Kluger and DeNisi (1996) feedback meta-analysis, via secondary glossary; check original.
• 10. Meyer, Kay and French, GE studies (1965), via secondary glossary; check original.
• 11. Gartner HR survey (March 2026), gartner.com press release.
• 12. Future of Privacy Forum, generative AI considerations (June 2024).
• 13. Ethisphere, AI policy and governance guidance (2026), vendor advisory.
• 14. AMA survey (October 2025, 1,300+ workers), via HR Dive; check original.
• 15. EU AI Act, Articles 4 and 50; check EUR-Lex.

Not verified for this course: the Gallup 2018 and 2020 burn-out figures (original numbers not pulled), the Kluger and DeNisi figures (via a glossary), the Gallup quick-connect durations (primary page not located), the CHRO roundtable figure (secondary, unverified), the 90% talk-time figure (vendor, not empirical), the 2026 Emerald coaching study (results not summarised), and the August 2026 privacy framework update (not read). Monitoring-disclosure guidance has no dated source.