An employee survey spends goodwill before it earns any. People answer honestly once. If nothing visible happens, the next round gets the answers they think you want.
Gallup says "23% of Indian employees were engaged in 2025 — the lowest point in four years — while 18% were actively disengaged (based on three-year rolling averages)." Plenty of those firms run surveys.
Should you run a survey at all?
Only if you can act. If the likely answers point at pay, workload or a specific manager, and you cannot move any of those, a survey makes things worse by naming a problem you then ignore.
Ask yourself what you would do if 40% said the same negative thing. If the honest answer is nothing, do not ask.
How short should it be?
Short enough to finish on a phone in 3 minutes. 10 to 12 questions, mostly scaled, with 1 or 2 open boxes.
Long surveys get lower completion and worse answers, because people stop reading carefully around question 20.
What should you actually ask?
| Area | A question that works |
|---|---|
| Expectations | Do you know what is expected of you this quarter? |
| Manager | Have you had a useful conversation with your manager in the last month? |
| Fit | Does your work use what you are good at? |
| Follow-through | When something is raised here, does it get fixed? |
| Intent | Do you expect to be here in a year? |
The fourth question predicts the value of the survey itself. If people already say nothing gets fixed, start there rather than asking more questions.
Is anonymity real?
Only if the reporting cannot identify anyone. In a 1,200-person firm that is easy. In a 12-person office it is not. A team of 4 filtered by tenure and location is not anonymous, whatever the tool claims.
Set a minimum group size for reporting, say 6, and say so up front. People calibrate their honesty to how exposed they feel.
What happens after the results?
Three steps, in order. Publish what you heard, including the uncomfortable parts. Pick 2 things to fix. Say what you are not fixing and why.
The third step is the one firms skip, and it is the one that buys credibility. People accept a no with a reason far better than silence.
Where do surveys hide a basic failure?
Behind morale language. A team scoring low on trust may simply be the team whose payroll was wrong twice, or whose benefit enrolment was never filed.
The Employees State Insurance Corporation says "ESIC provides reasonable Medical Care for self and family from day one of entering into insurable employment." An employee who found out the hard way that their cover was missing is not telling you about culture.
How often should you ask?
Annually for the full survey, with 1 or 2 pulse questions each quarter. India sat at 23% engaged in 2025 against 21% for South Asia and 20% globally, so expect movement in points, not leaps. More often than that and you are measuring faster than you can act.
Gallup says "Twenty percent of employees globally are engaged, according to the State of the Global Workplace: 2026 Report", so a flat result is not unusual. Treat a stable score with visible fixes as better than a rising score with none.
What should you measure besides the score?
Completion rate, which tells you whether people think it matters. Issues closed from the last round. And attrition inside 90 days by manager, which moves before any survey does.
PF and ESI records sit with the Employees Provident Fund Organisation, and the wider framework with the Ministry of Labour and Employment. Getting the basics right underpins everything on 247HRM engagement.