People leave for many reasons: pay, working conditions, life changes, contract endings or restructuring. An employee survey cannot reliably tell you who will resign. It can help teams discuss their experience and identify conditions worth improving.
Calculate turnover consistently
Turnover (%) = departures during the period / average headcount during the same period × 100. Define the population, period and method for average headcount. A simple average of opening and closing headcount may be inadequate when staffing changes substantially.
Six departures over a year with average headcount of 60 equals 10% annual turnover. Do not directly compare it with 10% in one quarter. Separate voluntary departures, dismissals and contract endings. Define any “regretted departure” category before reviewing the results.
Add context, not individual risk labels
Compare HR records with aggregate survey findings, exit interview themes and onboarding feedback. Examine periods and cohorts while protecting small groups from identification. Country comparisons also need context about employment arrangements and labour markets.
Gallup’s research links engagement with turnover at business-unit level. That is not a validated prediction of an individual’s resignation, or evidence that our platform reduces turnover.
Never try to match anonymous comments to individuals. A low eNPS answer, a self–manager rating gap or a negative comment is not a diagnosis of resignation risk.
Choose actions you can test
- Discuss workload, pay, work organisation and support with the team.
- Compare recruitment promises with everyday experience; use an onboarding plan.
- Agree realistic development opportunities without promising everyone a promotion.
- Improve one-on-one conversations and communication about decisions.
- Assign action owners and dates to review implementation and outcomes.
These are options to investigate. Recognition cannot substitute for fair pay, and training cannot resolve every organisational problem.
Estimate costs and review change
Use your own recruitment, vacancy, onboarding and team workload costs. Avoid one salary multiplier for every role. Compare turnover over consistent periods and examine workforce changes. A lower rate alone does not prove that an intervention worked. See interpreting results.
Start with organised feedback
The platform helps collect feedback and organise follow-up. Decisions still require conversation and context. Create an account and prepare an employee experience survey.
Sources and scope
These sources support the principles discussed here. Examples and schedules are starting points to adapt, not standards or evidence that our templates have been validated.
