The metrics most companies should track about their own teams

01.1. Regretted attrition

Total attrition is a blunt figure. Losing a person you were glad to see go is different from losing a strong performer you wanted to keep. Regretted attrition counts only the exits you would have prevented if you could.

To track it, agree a simple rule for marking an exit as regretted, such as performance and criticality, and apply it consistently. Look at it by team and tenure. A rise in one team is a signal to ask why.

02.2. Time to hire and time to productivity

Time to hire measures how long a role stays open, from approval to accepted offer. Long delays cost output and add pressure on the team.

Time to productivity measures how long it takes a new joiner to work at full speed. It shows whether onboarding works. Define a milestone that marks productivity for each role, such as the first independent release or the first closed case, and track the time to reach it.

Together, the two show the real cost of a hire.

03.3. Internal mobility

How many roles are filled by people already in the company, and how many employees move into new roles each year? Healthy mobility keeps knowledge in the business and gives people reasons to stay.

Low mobility may mean people do not see opportunities or that managers hold on to talent. Review it by function, and ask what blocks moves.

04.4. Manager span and team workload

Span is the number of people a manager leads. Very wide spans mean little coaching. Very narrow spans can mean too many layers. Neither is wrong everywhere, but an outlier deserves a look.

Workload is harder to measure, and imperfect proxies are fine: overtime patterns, leave taken, open tasks per person. Where the workload is unevenly spread, the risk of burnout and exits is higher.

05.5. Engagement trend by team

A single engagement score is less useful than its direction. Track it over time by team, using short, regular pulse questions rather than one long annual survey.

Watch the trend and the response rate. A falling response rate is itself a signal. Use minimum group sizes so no individual's answers can be identified, and act visibly on what people say. If nothing changes after a survey, people stop answering.

06.6. Pay equity gaps

Compare pay for comparable roles, levels and locations across groups. Gaps are not always unfair, as there can be good reasons, but they should be explained.

Handle this metric carefully. Restrict access to a small group, work with legal advice, and use the findings to review decisions, not to single out people.

07.A worked example: three metrics for a growing company

A company of 400 people has hired fast for a year. Leadership picks three metrics: time to productivity, regretted attrition and manager span. The first review shows time to productivity is long in two functions, regretted attrition is rising in one team, and several managers have more than fifteen reports.

Those three facts lead to three actions. The company improves onboarding in the slow functions. It talks to the team with rising exits and finds a role change nobody explained. It splits the widest teams. Nothing here needs complex tooling: a clean data set and a monthly meeting were enough.

08.Where the data comes from

Most of these metrics can be built from data you already hold: HR records for headcount, joiners and leavers; the applicant tracker for time to hire; pulse tools for engagement; payroll for pay. The challenge is joining them with consistent employee identifiers and agreed definitions.

Check the basics first. Are leaving dates recorded reliably? Do job titles map to levels? Is the manager field current? Small data fixes at the source remove hours of manual work each month.

09.Reading the numbers with care

A metric is a prompt for a question. A rise in attrition in a team of eight may be one resignation. A fall in engagement may follow a reorganisation. Look at trends over several periods, compare similar teams, and talk to people before drawing conclusions.

Beware targets that distort behaviour. If managers are judged on low attrition, they may discourage healthy exits or hide them. Use metrics to start discussion, not to rank people.

10.How to choose where to start

You do not need all six at once. Pick three that match your biggest concerns.

• If you are losing people, begin with regretted attrition, engagement trend and manager span.
• If you are growing quickly, begin with time to hire, time to productivity and internal mobility.
• If you suspect fairness issues, begin with pay equity, internal mobility and attrition by group.

Write a one-line definition for each metric. Decide who owns it. Review it monthly, in the same meeting, so trends are visible.

11.A note on definitions

Write each metric down in one sentence before you track it. For regretted attrition, say who decides and when. For time to hire, say which two dates you use. For span, say whether contractors count. Keep a short glossary where leaders can find it. When someone asks why a number moved, the first check should be whether the definition did.

12.Protecting privacy

Metrics about teams should not become surveillance of people. A few rules help.

• Report at team level, and hide groups below a minimum size.
• Limit who sees what, by role, and log access.
• Collect only what you need and say why.
• Tell employees what is measured and how it is used.
• Never use these metrics as the sole basis for a decision about an individual.

13.Common mistakes

• Tracking too much. Twenty metrics become noise.
• Changing definitions. If the meaning shifts, trends become meaningless.
• Looking only at averages. Averages hide the team in trouble.
• Measuring without acting. Every metric should link to a decision someone can make.
• Ignoring data quality. A scorecard built on inconsistent records misleads.

14.A monthly review routine

Hold a short monthly meeting with HR and business leaders. Show the three chosen metrics for the last six months, by team. Ask what changed, what the likely reasons are and who will look into each outlier. Record the actions, and check them next month. After a quarter, decide whether to keep, swap or add a metric.

The routine matters more than the dashboard. A scorecard nobody discusses changes nothing, and a simple one discussed every month changes a lot.

15.What to do next

Pick your three metrics this week. Write the definitions. Find where the data lives and how clean it is. Build a first scorecard, even a simple one, and review it at next month's leadership meeting.

Measuring people well is a matter of care more than technology, but good tooling makes it practical. At Brandsmashers we build people-analytics dashboards that bring data from several HR systems into one reviewed view, with privacy and access controls designed in. If your data sits in too many places, a scoping call can help you plan a clean start.