How do you measure employee engagement? Metrics, scores and methods that hold up
Ask three providers how engaged your people are and you can get three numbers that cannot be compared: an average out of five, the share of people who agreed with a set of statements, and the share who cleared a much higher bar. Each is called an engagement score. The number on your dashboard is only as meaningful as the method behind it.
The short answer: measure engagement three ways and look for agreement. Ask people (a survey, ideally one that makes them choose between things they value). Watch what people do (voluntary turnover, absence, internal moves, the extra effort nobody requires). Talk to people (stay interviews and focus groups). Report results by group, not only as one average, and judge a score by its method before you judge it by its size.
What exactly are you measuring when you measure engagement?
Engagement is how much of themselves people bring to their work — not how content they are with the conditions around it. Kahn's original definition (1990) describes people employing and expressing themselves physically, cognitively and emotionally in their work roles. The most widely used research scale describes work engagement as a positive state characterised by vigour, dedication and absorption (Schaufeli, Bakker and Salanova, 2006).
Satisfaction is a different thing. A person can be satisfied with pay, hours and a calm office, and still give the minimum. Another can be frustrated with the same conditions and fully committed to the work. Most "engagement surveys" measure a mix of both, which is why the label on the report matters less than the questions inside it.
Method changes the headline dramatically. Gallup counts an employee as engaged only when their answers clear a high bar; Gallup itself warns that this is not comparable with "percent favourable" indices that combine "agree" and "strongly agree". By Gallup's measure (141,444 employed respondents in more than 140 countries), 20% of employees worldwide were engaged in 2025. In its three-year country averages to 2025, the figure was 4% in Egypt, 26% in Saudi Arabia and 27% in the UAE, each with a margin of up to about seven points. Those are national pictures under one strict method — not benchmarks for a company score built another way.
Which methods measure employee engagement?
No single method sees everything. The table sets out what each one captures and what it misses.
| Method | What it captures | Strength | Blind spot | Why use it |
|---|---|---|---|---|
| Annual engagement survey (rating scales) | How people judge many aspects of work, at one moment | Broad coverage; comparable year on year if nothing changes | Picks up the mood of the weeks around fieldwork | A full baseline across the organisation |
| Pulse survey | A few questions, repeated often | Shows direction between big surveys | Too few items to explain why something moved | Checking whether an action worked |
| Forced-choice survey | What people prioritise, and what they feel they receive | Less exposed to mood and to agreeing by habit | Needs careful design and scoring | Finding the gap between what people want and what they get |
| Stay interviews | Why people stay, and what would make them leave | Depth, and names the specific trigger | Small numbers; depends on trust in the interviewer | Retaining people before they resign |
| Focus groups | Shared views and the language people use | Explains a survey result | Louder voices dominate without facilitation | Understanding a number before acting on it |
| Exit interviews | Why people left | Leavers may speak more freely than current staff | Arrives too late for that person; leavers often soften criticism to protect a reference | Patterns across leavers |
| Behavioural data | What people do, not what they say | Recorded anyway, so nobody answers it for effect | Many causes behind each signal | Testing whether survey results show up in reality |
Combine at least one method that asks, one that watches and one that talks. When a survey result, a behavioural trend and an interview theme agree, act on it; when they disagree, find out why first.
How do you calculate an employee engagement score?
There are four common ways, and each can mislead in its own way. Choose one, state it on every report, and never compare scores calculated differently.
| Method | How it is calculated | What it hides | Why it can mislead |
|---|---|---|---|
| Average score | Mean of all engagement items on a 1–5 scale | The spread: a team split into very happy and very unhappy averages to "fine" | Small changes look precise when they are within normal noise |
| Percent favourable | Share of answers that are "agree" or "strongly agree" across the items | How strongly people agree, and what the "neutral" group thinks | Moves a lot when people shift between "neutral" and "agree" |
| Index of engaged people | Share of people whose answers clear a defined bar on every item or on average | How close everyone else is to the bar | Depends entirely on where the bar is set |
| Employee Net Promoter Score (eNPS) | Share answering 9–10 on "how likely are you to recommend us as a place to work?" minus the share answering 0–6 | Everything except willingness to recommend | One question, very sensitive to mood and to who answers |
For example, if 120 of 200 answers are 4 or 5, percent favourable is 60%.
Whatever the method, report three things with every score: the response rate, the number of people behind each group's result, and the distribution — not just the average. And set a minimum group size before anything is reported. The GK Engagement Survey never reports a group of fewer than seven people.
How can you measure engagement without a survey?
Through behaviour you already record — not through another questionnaire. None of these is a measure of engagement on its own; together they show whether what people say matches what they do.
| Behavioural signal | How to calculate it | What it can signal | Caveat |
|---|---|---|---|
| Voluntary turnover rate | Voluntary leavers in a period divided by average headcount, times 100 | People choosing to leave | Labour-market conditions move it too; our turnover calculator shows the working |
| First-year turnover | Hires in a period who left within 12 months, divided by all hires in that period | Whether the job matched the promise | Small cohorts swing widely |
| Unplanned absence | Unplanned absence days divided by scheduled working days | Withdrawal before resignation | Health and family events dominate in some months |
| Internal mobility | Internal moves and promotions divided by headcount | Whether people see a future inside | Depends on how many roles exist |
| Discretionary participation | Share taking part in optional programmes, suggestions or referrals | Effort nobody requires | Can be inflated by pressure to take part |
| Employee referrals | Hires from employee referrals as a share of all hires | Willingness to stake one's own name on the employer | Referral bonuses change the motive |
Which employee engagement KPIs should you report to leadership?
Five are enough for most leadership teams, and each one leads to a decision — not just a slide.
- Participation rate, by group. Low participation is a result in itself: it often means people doubt that anything will change.
- Engagement result, by group and with its method stated, compared only with the same method.
- Top priority and top frustration, by group: what people want most, and where what they get falls furthest short.
- Voluntary turnover in the first year, by group, as the behavioural check.
- Action completion: the share of teams with an agreed action, a named owner and a date, and how many of those actions were delivered.
The last KPI is the one most often missing. In Gallup's 2024 research, only 8% of employees strongly agreed that their organisation takes action on surveys, and 38% said they did not know whether it did. Gartner's 2023 survey of nearly 3,500 employees found that only about a third believed their organisation would act on their feedback.
Why does measurement fail after the survey?
Because the report arrives, the presentation goes well, and nothing visible happens. The next survey then measures disappointment as well as engagement.
Set the follow-through rules before the survey launches, not after the results land. In the GK Engagement Survey the timetable is fixed in advance: a sponsor report within 14 days of the survey closing, the headline to everyone within 30, and one action per group within 45, each with an owner and a date. If you already run your own survey, our earlier article on engagement surveys that move the needle covers how to close the loop with managers.
And if this year's results came back lower than the year deserved, read why engagement scores drop in a hard year before you present them.
Frequently asked questions
What is a good employee engagement score?
There is no universal number, because scores depend on the method. A good score is one that is higher than your own last result measured the same way, and whose gaps between groups are shrinking. Compare with external benchmarks only when they use exactly your method.
How often should you measure employee engagement?
A full survey once or twice a year, short pulse surveys between rounds to check whether actions worked, and behavioural data every month. Measuring more often than you can act teaches people that surveys change nothing.
What is the difference between employee engagement and employee satisfaction?
Satisfaction is contentment with the conditions of work: pay, hours, environment. Engagement is how much of themselves people put into the work. You need both, but they are not the same, and improving one does not guarantee the other.
Is eNPS a good measure of employee engagement?
It is a useful single signal of advocacy, and quick to run. It is not a full measure of engagement: one question cannot tell you why people answered as they did, and it moves with mood.
Which tool is best for measuring employee engagement?
The one whose method you can explain, that people complete in their own language, and that ends in actions with owners and dates. Ask any provider how the score is calculated, what the minimum reporting group is, and what happens in the 45 days after the survey closes.
Sources
- Kahn, W.A. (1990), "Psychological conditions of personal engagement and disengagement at work", Academy of Management Journal 33(4). doi.org
- Schaufeli, W.B., Bakker, A.B. and Salanova, M. (2006), "The measurement of work engagement with a short questionnaire", Educational and Psychological Measurement 66(4). doi.org
- Gallup, State of the Global Workplace 2026 (2025 data; 141,444 employed respondents in more than 140 countries) and country pages for Egypt, Saudi Arabia and the UAE (three-year averages to 2025). gallup.com
- Gallup (2024), "Action plans propel engagement". gallup.com
- Gartner (2023), press release on its survey of nearly 3,500 employees. gartner.com
- Bain & Company, Net Promoter System: how the score is calculated. netpromotersystem.com