Many companies track their managers by counting 1:1s held and training hours logged, then call the dashboard done. Those numbers look busy and rarely change a decision. That’s a problem, because managers carry a heavy share of a team’s results and the role is under strain. According to Gallup’s 2026 workplace report, manager engagement fell to 22% in 2025, down from 31% in 2022, while global engagement sat at 20%. That pressure shows up in the data you’re about to collect.
This guide shows how to measure manager effectiveness with numbers that hold up when someone asks what they mean. It starts with a simple test for spotting vanity metrics, then builds a core set of manager effectiveness metrics that work in any function. Later sections deal with the problems that distort manager data, such as small teams and inherited issues. You’ll also find role-specific measures for SaaS teams and a step-by-step scorecard build. The goal is a short list of numbers you’d act on.
TL;DR
A practical way to measure manager effectiveness is to track a short list of numbers that would change a decision if they moved. Start with regretted attrition against the company baseline and upward feedback favorability. Then check whether each manager’s reports get meaningful feedback every week, and add 1 or 2 function measures, such as quota attainment spread for sales or DORA metrics for engineering. Read every number as a trend over several cycles and adjust for team size before you judge anyone. Keep developmental feedback away from pay. Put the final set on a manager scorecard and audit it every year.
What manager effectiveness means
Manager effectiveness is the difference a manager makes to a team’s results and health, compared with what the same team would likely do under an average manager. That comparison matters. A strong team can post good numbers under a weak manager for a while, and a strong manager can inherit a team in trouble. So you’re measuring contribution, which is harder than reading a single team score. A practical definition works in layers. Behavior covers what a manager does each week, like coaching and setting priorities. Outcomes cover team health, which shows up in retention and engagement, along with the team’s output in its function.
Behavior is where much of the research started. Google’s Project Oxygen research drew on more than 10,000 data points from reviews and surveys and landed on 10 behaviors of effective managers. They include being a good coach and supporting career development. Those behaviors are useful because a manager can change them. Outcomes still matter more for judging impact. The CIPD’s 2023 evidence review on people managers notes that a systematic review found most leadership assessment tools lack rigor. It advises measuring outcomes linked to people management when judging impact. When you measure manager effectiveness, behaviors tell you where to coach and outcomes tell you whether the coaching worked.
Why manager quality drives retention and output
Managers shape how engaged their teams are. In Gallup’s engagement variance analysis, the firm estimated that managers account for at least 70% of the variance in engagement scores across business units. That estimate comes from Gallup’s own data, so treat it as a strong signal from a large sample. Engagement also tracks with money. Gallup’s 2024 Q12 meta-analysis covered 183,806 work units and found that top-quartile teams had 23% higher profitability and 78% lower absenteeism than bottom-quartile teams. Those findings are correlations. They show the stakes, though they don’t prove a manager caused the gap.
Stronger evidence comes from economists who tracked output worker by worker. A study titled The Value of Bosses used productivity data from a large company’s technology-based service workers. It found that replacing a boss from the bottom 10% with one from the top 10% added more to team output than adding 1 worker to a 9-person team. Workers with better bosses were also less likely to quit. The authors concluded that most of a good boss’s value comes from teaching, which means workers build skills that stick. That’s a useful lens for measurement, because teaching pays off slowly.
The clearest link to retention comes from a Hoffman and Tadelis study of a large high-tech firm, published in the Journal of Political Economy in 2021. The researchers used a 6-question upward survey where employees rated their managers. Moving from a 10th-percentile manager to a 90th-percentile manager on people skills was associated with about a 60% drop in turnover. The effect was strongest for regretted quits, the people the company most wanted to keep. The authors estimated that lower hiring costs alone cut labor costs for the manager’s team by about 5%. That’s why the numbers you use to measure manager effectiveness should tie to these outcomes.
The vanity test for manager effectiveness metrics
A vanity metric is a number that looks good on a slide and doesn’t change what anyone does. Several common manager effectiveness metrics fall into this trap because they count activity. A manager can hold every scheduled 1:1 and still leave reports confused about priorities. Before a metric earns a spot on your dashboard, run it through 4 questions. First, would a different reading change a decision about coaching, support, a promotion or a role change? Second, can the manager influence it within 1 or 2 quarters? Third, is it measured on enough people to separate a real signal from noise? Fourth, is it hard to hit without doing the actual work?
A metric that fails 2 or more of these questions is decoration. Vanity metrics still have uses, such as checking that a program is running. They just can’t carry a judgment about a manager. The table below shows common vanity picks and the decision metrics that can replace them. Each replacement asks a harder question about the same area.
| Vanity metric | Why it misleads | Better decision metric |
| Number of 1:1s held | Counts meetings and says nothing about their value | Share of reports who got meaningful feedback in the past week |
| Manager training hours | Measures an input that may never change behavior | Change in upward feedback favorability across 2 or more cycles |
| Team headcount growth | Reflects budget decisions made above the manager | Regretted attrition compared with the company baseline |
| Team total revenue | Often driven by territory or an inherited book | Spread of quota attainment across reps |
| Company-wide eNPS | Can’t be traced to any single manager | Team-level favorability on manager items, with 5 or more respondents |
| Promotions granted | Can reflect leniency more than growth | Performance of promoted reports 12 months later |
Core manager effectiveness KPIs for every team
A small core set of manager effectiveness KPIs works in almost any function, from sales to engineering. These 4 cover team health from different angles, and you can calculate each one with data your HRIS and survey tool usually capture. Together they let you measure manager effectiveness without leaning on activity counts.
Regretted attrition compared with the company baseline
Regretted attrition is the clearest people outcome to track. Count the voluntary exits of people the company wanted to keep over the trailing 12 months, then divide by the team’s average headcount. Compare that rate with the company rate for the same period. Raw attrition misleads because it lumps in healthy exits, and the Hoffman and Tadelis data found that good managers mainly cut regretted quits, with no drop in the others. The cost of getting this wrong is high. Gallup estimated in 2019 that replacing an employee costs between 50% and 200% of their annual salary, and 52% of voluntary leavers said their manager or organization could have prevented the exit.
Upward feedback favorability
Favorability is the share of reports who answer 4 or 5 on a 5-point scale for each manager item. It comes from an upward feedback survey, where reports rate specific behaviors such as clarity and coaching. Track the average across items and the trend on the 2 or 3 items that matter most for that team. Some companies also add a team-level eNPS. Bain’s formula subtracts the share of detractors, who score 0 to 6, from the share of promoters, who score 9 or 10. It’s a blunt tool for judging 1 manager, so keep it as context. Survey design gets its own section below.
Meaningful feedback each week
Ask reports directly whether they got meaningful feedback in the past week, and track the share who say yes. Gallup’s meaningful feedback research found that 80% of employees who received meaningful feedback in the past week were fully engaged. The same work found that only 16% of employees described their last conversation with their manager as extremely meaningful, so most teams have room to improve. Gallup also found that frequent conversations of 15 to 30 minutes tend to beat occasional long ones. A survey item captures quality in a way a calendar count can’t.
Internal mobility from the team
Good managers grow people, and some of that growth means people leave the team for better internal roles. Track the share of a manager’s reports who were promoted or moved to another internal role over 12 to 24 months. Lateral moves count too. MIT Sloan attrition research by Donald Sull and colleagues found that lateral career moves were 12 times more predictive of retention than promotions. The same 2022 study found that a toxic culture was 10.4 times more powerful than pay in predicting attrition. A manager who sends talent to other teams is often doing the job well, even if their own team churns.
How to design an upward feedback survey managers trust
An upward feedback survey only works if reports answer honestly and managers believe the results. Google’s version is a well-documented example. Google’s upward feedback program asked employees to rate managers on a 5-point scale every 6 months, starting in June 2010, and the company timed it away from performance reviews on purpose. Results stayed developmental. Between 2010 and 2012, median favorability rose from 83% to 88%, and the lowest-scoring managers improved the most. Each item mapped to a specific behavior, which gave managers something concrete to work on. Run well, this kind of survey is one of the most direct ways to measure manager effectiveness.
Keep the survey short, with 6 to 12 items scored on a 5-point scale. Add 2 open questions about what the manager should keep doing and what they should change. Every item should describe something a report can observe. Sample items might read like this:
- My manager gives me feedback I can act on.
- My manager supports my career growth.
- My manager sets clear priorities for the team.
- My manager lets me do my work without micromanaging.
- My manager recognizes good work.
- My manager treats people on the team fairly.
Anonymity rules decide whether people answer honestly. UC Davis shows supervisors their scores only when at least 5 people respond and their comments only when at least 10 do. The US federal employee survey reports only units with 10 or more respondents. The high-tech firm in the Hoffman and Tadelis study set its minimum at 3 reports in the first year and raised it to 5, and its survey drew about a 95% response rate. For small teams, roll upward feedback survey results up to the next level. Watch the response rate too, since a sudden drop on 1 team can say more than the scores do.
Why small teams produce noisy manager data
Small teams produce the most extreme scores at both ends, and that fact can break manager rankings. Statistician Howard Wainer called the link between sample size and variation the most dangerous equation because ignoring it leads to so many bad decisions. His examples include Pennsylvania schools, where the smallest schools showed up far more often than expected in both the top 50 and the bottom 50 on test scores. Managers work the same way. A team of 5 can swing from top to bottom in a single survey cycle without anything real changing.
A single exit or a single survey wave says little on its own. Take a hypothetical 6-person team. If 1 person the company wanted to keep leaves, annual regretted attrition jumps to about 17%, which looks like a crisis next to a company rate in single digits. The exit might have had nothing to do with the manager. Single readings are noisy even on larger teams. In the Hoffman and Tadelis data, a manager’s score in 1 survey wave correlated only about 0.37 with the next.
You can still measure manager effectiveness on small teams if you change how you read the numbers. Look at 3 or more cycles before drawing conclusions, and use rolling 12-month windows for attrition. Compare each manager with their own baseline. Show results as ranges when you can. Avoid ranking managers using raw scores from very small teams, especially those with fewer than 5 people, because 1 response can shift the result sharply. A reporting threshold protects anonymity, but it doesn’t guarantee that a sample is statistically reliable. Apply that caution to every upward feedback survey score you report.
How to adjust for context outside the manager’s control
Managers inherit situations they didn’t create, and fair measurement accounts for that. A new manager who takes over a team mid-reorg will often show weaker numbers than a peer who’s led a stable team for 3 years. A sales manager with a strong territory can post great totals with average coaching. Fair comparisons let you measure manager effectiveness without punishing bad luck. Your manager effectiveness KPIs should carry flags for these factors so a reviewer sees them next to the score. The table below covers the most common confounds and how to handle each one.
| Context factor | How it distorts the reading | How to adjust |
| Inherited team | Past problems show up in the new manager’s numbers | Set a baseline at handover and judge the trend from there |
| Territory or segment | Strong or weak accounts drive results more than coaching | Compare only with peers in the same segment |
| Span of control | Large teams dilute attention and small teams add noise | Group managers by span band before comparing |
| Reorgs and manager changes | Teams in flux often dip on engagement and retention | Flag the affected quarters and reset baselines |
| Time in the role | First-year managers are still learning the job | Compare new managers with a new-manager cohort |
| Company-wide events | Layoffs or pay changes move every team at once | Measure each team against the company-wide change |
Manager capacity is a hidden variable
A manager with no time for people work will score poorly on people metrics, and the cause may sit in the org design. A McKinsey middle manager survey published in 2023 found that middle managers spend nearly half their time on non-managerial work, including almost a full day each week on administration. Less than 1/3 of their time went to managing talent and people. Deloitte’s 2025 Human Capital Trends research found a similar split, with managers spending nearly 40% of their time on present-day problems and admin and only 13% on developing their people.
The CIPD suggests that non-managerial work should take up no more than 30% of a manager’s time. Gartner reported in 2024 that 75% of HR leaders say their managers are overwhelmed by growing responsibilities. So track capacity next to performance. A quarterly calendar sample or a short self-report on the share of time spent on people work is enough. Any attempt to measure manager effectiveness should account for this, so put that figure on the manager scorecard as context. If a manager’s upward scores fall while their individual workload climbs, the first fix is the workload.
Function-specific manager metrics in SaaS teams
The core set covers team health, but each SaaS function also needs 1 or 2 output measures. These manager effectiveness KPIs should reflect what the manager controls inside that function, with the same care around context and sample size covered earlier.
Sales managers
Track how quota attainment is spread across reps. A team can hit its number because 1 star rep carried it or because the manager closed deals personally, and neither tells you much about coaching. Look at the share of reps at or above quota and how fast new reps ramp compared with the company’s past hiring cohorts. Don’t judge sales managers on their own selling record. A Peter Principle sales study of 214 firms found top sellers were more likely to be promoted, yet doubling a new manager’s pre-promotion sales was linked to a 7.5% drop in their team’s sales. Public benchmarks for attainment and ramp come mostly from vendors, so use your own history as the baseline.
Customer success managers
Gross revenue retention (GRR) is usually the cleaner measure for a CS manager’s book. GRR counts revenue kept after churn and downgrades and leaves out expansion. Net revenue retention (NRR) adds expansion, which sales or account management often shares. In its 2021 cloud portfolio analysis, Bessemer Venture Partners reported median NRR of 125% for firms with $1 million to $10 million in annual recurring revenue and 115% for those above $100 million. Gross retention was generally 85% to 90% across revenue bands. These figures came from Bessemer’s portfolio data covering 2010 through the first half of 2021, so they aren’t universal benchmarks; SMB-focused businesses may run lower. Compare a CS manager’s GRR trend with peers who manage a similar segment, and watch for noise in small books where 1 lost logo moves the whole number. Pair the retention number with CSM regretted attrition so retention isn’t bought with burnout.
Engineering managers
Engineering managers can use DORA’s 5 delivery metrics as the team output layer. They’re change lead time, deployment frequency, failed deployment recovery time, change fail rate and deployment rework rate. The first 3 track throughput and the last 2 track instability. The DORA team warns against turning these numbers into targets or comparing teams that build very different software. Use them to follow a team’s trend over time. Pair them with upward feedback and retention, since a team can ship faster for a quarter while its people burn out. Google’s Project Aristotle studied 180 teams and found psychological safety was the strongest of the 5 team dynamics it measured.
How to keep manager metrics from being gamed
Numbers tied to rewards tend to get gamed. Anthropologist Marilyn Strathern summed up Goodhart’s law in 1997: “When a measure becomes a target, it ceases to be a good measure.” An HBR surrogation article by Michael Harris and Bill Tayler describes how people start treating the metric as the goal itself. At Wells Fargo, the push for cross-selling led staff to open about 3.5 million accounts without customer consent. The authors advise using several metrics and avoiding incentives tied to any single one.
Manager effectiveness metrics are especially exposed because managers can influence how reports answer surveys and when people exit. Google handled this by keeping upward feedback developmental and out of performance reviews. You can follow the same split. Let the upward feedback survey guide coaching, and let pay decisions rest on outcome measures read in context. Before any manager scorecard goes live, check it against this list:
- No single metric decides pay or promotion
- Upward feedback results stay developmental
- Every activity measure is paired with an outcome measure
- Rules on who sees which data are written down
- Sudden jumps in a score trigger a closer look
- Survey items are reviewed and refreshed each year
- Managers know why each metric exists
How to build a manager scorecard step by step
A manager scorecard pulls the earlier sections into a single view that a manager and their leader review together each quarter. Keep it short enough to discuss in 30 minutes, and build it in this order.
- Name the decisions it will inform. List what the scorecard should help you decide, such as who needs coaching and who’s ready for a bigger team. If a metric can’t feed one of those decisions, leave it out.
- Pick 4 to 6 metrics that pass the vanity test. Make sure the set covers behavior as well as outcomes. A sensible starting set is regretted attrition, upward feedback favorability, weekly meaningful feedback and 1 function measure.
- Set minimum sample sizes. Report survey scores only with 5 or more respondents and comments only with 10 or more. Roll smaller teams up a level and use 12-month windows for attrition.
- Define comparison groups and context flags. Compare managers within the same function and span band. Add flags for inherited teams and reorgs so nobody reads a score without its context.
- Split developmental and evaluative views. Managers should see all of their own data. Their leaders should see trends and outcome measures, while detailed survey comments stay with the manager and HR.
- Set the cadence. Review the manager effectiveness KPIs quarterly and run the upward survey every 6 months. Read attrition on a rolling 12-month basis so a single quarter doesn’t swing the result.
- Review trends and audit yearly. Walk through trends with each manager and agree on 1 or 2 behaviors to work on. Every year, audit the scorecard itself and drop any metric that never changed a decision.
What changes for manager measurement in 2026
Teams are getting wider, and that changes what manager data can tell you. A Gallup span analysis published in January 2026 shows the average US manager had 12.1 direct reports in 2025, up from 8.2 in 2013. The median stayed near 6, because a few very large teams pull the average up. In an October 2024 forecast, Gartner predicted that through 2026, 20% of organizations would use AI to flatten their structure and eliminate more than half of their middle management roles. That’s a prediction, not a measured 2026 result. If spans widen and teams reorganize, baselines may need resetting more often.
AI adds a new behavior to measure. The Microsoft 2026 Work Trend Index cites a separate study of 1,800 workers. Employees whose managers modeled AI use reported 17-point higher AI value and 30-point higher trust in agentic AI. These are survey associations, not proof that manager behavior caused the differences. Consider adding an upward item about whether the manager helps the team use AI well. The labor market matters too. The Bureau of Labor Statistics reported a quits rate of 1.9% for August 2026. When fewer people quit, attrition becomes a slow signal, so leading indicators deserve more weight as you measure manager effectiveness this year.
Where to start with manager measurement
Before rolling out a manager scorecard, write down the decision it should support. You might need to identify coaching needs, spot team-level risks or check whether managers have enough time for people work. Choose the question first, then use the metrics that could change your response. That makes the first review about a real management choice rather than another dashboard.
At the review, note where the numbers disagree. A manager with steady output but falling upward feedback may need support before turnover rises. A manager with higher attrition after a reorg may need time and a revised baseline. Record the action you take and revisit it next quarter. The scorecard earns its place when it helps you make a fairer decision.
Frequently asked questions
What is a good manager effectiveness score?
There’s no universal benchmark because survey items and scales differ between companies. Compare a manager with their own earlier scores and with peers in the same function. Treat scores from small teams as a prompt for discussion rather than a precise ranking.
How often should you measure manager effectiveness?
Review the scorecard quarterly and run the upward survey every 6 months. Read attrition on a rolling 12-month window. Weekly feedback checks can run continuously as part of normal 1:1s. Avoid monthly team-level pulses on small teams, since the results swing too much to act on and survey fatigue lowers response rates over time.
Should manager effectiveness scores affect pay?
Keep upward feedback scores out of pay formulas. Once a survey result drives compensation, managers have a reason to influence how reports answer, and honest feedback dries up. Pay and promotion decisions can draw on outcome measures such as regretted attrition and team results, reviewed with context. Keeping the 2 uses apart protects the honesty of the survey.
What is the minimum team size for reliable manager data?
A threshold such as 5 respondents can help protect anonymity, but it doesn’t make a small sample statistically stable. Some organizations require 10 responses before sharing comments. For very small teams, combine groups or look across several cycles before drawing conclusions.
Can you measure manager effectiveness without an HR platform?
Yes. Small companies can measure manager effectiveness with HRIS exports and a basic survey tool. Regretted attrition needs exit dates and a simple regretted tag. Upward feedback needs a confidential survey with a clear anonymity rule. The manual work is manageable until you have dozens of managers, and the logic stays the same after you add software.
How do you evaluate a new manager in the first 90 days?
Treat the first 90 days as a baseline period. Outcome measures like attrition reflect the previous manager more than the new one at this stage. Use a short set of behavior questions at around 60 days, along with regular check-ins between the new manager and their leader. Judge outcomes against that baseline after 2 full survey cycles.
How is manager effectiveness different from leadership effectiveness?
Manager effectiveness looks at how well someone runs a specific team, including clarity, coaching, retention and output. Leadership effectiveness covers a wider scope, such as setting direction and influencing people across the organization. Senior leaders need both. For front-line and middle managers, team-level measures give the clearer signal because the manager’s influence is direct and easier to observe.
Is eNPS a fair way to judge a single manager?
Rarely. Employee NPS asks whether people would recommend the company as a place to work, so company-wide factors like pay shape the answer. On a 5-person team, 1 person changing their answer can move the score by 20 points or more. Use team eNPS as context next to survey items that ask about the manager directly.
How do you measure managers of managers?
Look at the health of the teams below them and at how their managers develop. Useful measures include regretted attrition across the whole group and the upward scores of the managers who report to them. Skip-level survey items about clarity also help. Over time, check how many of their managers grow into larger roles.
What should a manager do after a low upward feedback survey score?
Start by sharing the main themes with the team and thanking them for honest answers. Pick 1 or 2 behaviors to work on and ask the team what better would look like. Check progress informally each month and look for movement at the next survey cycle. A low score from a single wave is a starting point for coaching.
How do you measure a player-coach who also carries a quota?
Separate their personal output from the team’s results. Use the same team measures you’d use for any other manager, and record the share of time they spend on people work. If their upward scores slip as personal quota pressure rises, the role design is the likely cause. Fix the workload before you judge the manager.
Does a 360 review replace an upward feedback survey?
No. A 360 review gathers input from peers and leaders as well as reports, and it’s often used for evaluation. An upward survey collects ratings only from a manager’s direct reports on specific management behaviors, and it works best when kept developmental. Some companies run both, with the upward survey on a regular cycle and 360s less often.
How do remote and hybrid teams change manager measurement?
The core measures stay the same, though proxies like online status or meeting counts become even less useful. Lean on survey items about clarity and feedback, since remote reports can’t pick up these signals in passing. Gartner found that hybrid managers were 2 times as likely as onsite managers to report having more direct reports, so check span when comparing scores.
Should managers see their own manager scorecard?
Yes. Managers should see their own results, including trends and comment themes that meet anonymity rules. Give peer comparisons as bands rather than exact rankings. That lets managers act on the findings without overreading small differences.
Which manager effectiveness KPIs matter most at an early-stage startup?
With only a few managers, keep it simple. The manager effectiveness KPIs that matter most early are regretted attrition and a short upward survey, backed by regular skip-level conversations. Founders can run skip-levels directly, which gives richer context than any score at this size. Add formal metrics as the manager count grows past 5 or 6.
