Measuring Employee Wellbeing with Surveys and Indicators

How do you measure employee wellbeing? Which surveys and indicators to use and how to read results: a practical guide for HR and OHS professionals.

In a management meeting, the HR manager is asked: "Did the support programme we launched last year work?" They have attendance lists and satisfaction forms from a few events, but no clear answer on whether absence, resignations or employees' stress levels have changed. This is very common, and it is one of the main reasons programmes are dropped at the first budget cut. Sound wellbeing measurement both shows you where to start and lets you track the results of your investment. In this article, for business owners, HR managers and occupational health and safety (OHS) specialists, I go step by step through which data to collect, how to design surveys and how to interpret the results.

Why Should Employee Wellbeing Be Measured?

Measurement has three core functions:

• Diagnosis: Seeing which risk factor stands out in which unit and on which shift. A general impression that "there is stress" is not enough for an action plan.
• Prioritisation: Directing limited budget and time to the areas that will have the greatest impact.
• Monitoring and accountability: Seeing whether the steps taken make a difference and reporting to senior management with concrete data.

Measurement also has a legal dimension. Article 8 of the Occupational Health and Safety Risk Assessment Regulation provides that psychosocial sources of hazard should be identified and recorded and, if they have not previously been examined, that the necessary controls, measurements and investigations should be carried out. The Ministry of Labour and Social Security's (ÇSGB) 2024 "Information Guide on Psychosocial Factors" also lists ten hazard categories such as workload, work pace, control, interpersonal relationships and organisational culture, and includes measurement tools in its annex. The guide is advisory in nature; however, it is a good starting point for structuring a measurement exercise.

In the context of Turkey, the importance of measurement is even clearer. According to Gallup's 2026 report, 66% of employees in Turkey say they spent much of the previous day under stress, while only 19% rate themselves as "thriving" (Gallup, 2026; 2025 data). National averages give an idea; but only your own data can show the picture in your organisation.

The Difference Between Measurement and Screening

There is a distinction that is often confused in organisational wellbeing work. Measurement is aimed at understanding working conditions and employees' overall state at group level. Screening, on the other hand, tries to detect possible mental disorders in individuals.

The WHO's 2022 guidelines on mental health at work made no recommendation for screening programmes in the workplace, on the grounds that benefits and harms are uncertain. In practice this means that your organisation's survey should not be a diagnostic tool that labels employees "at risk" or "not at risk". The aim is to assess the climate and conditions, not individuals. If there is a need for individual assessment, it should be done confidentially and with a specialist.

This distinction also determines the language of the survey. Instead of clinical questions such as "In the last two weeks, have you felt hopeless?", statements focused on working conditions such as "My workload can be completed within working hours" or "I know I can go to my manager when I am struggling" should be preferred. Even so, a serious difficulty may be expressed during the survey or in open-ended comments. For this possibility, it is good practice to give information on confidential support channels at the end of the survey and to remind people that 112 (emergency services) should be called in an emergency.

Which Indicators Are Used in Wellbeing Measurement?

A sound measurement system does not rely on a single survey; it brings together different data sources. It is useful to think of these indicators in two groups.

Lagging indicators: the consequences of the problem

These indicators show the traces left after the problem has emerged, and in most organisations they are already recorded:

• Absence and sick days: The total number of days, as well as the trend in short, frequently repeated sick notes.
• Turnover rate: Especially people who leave within the first year, and voluntary departures.
• Workplace accidents and near misses: Indirect indicators of fatigue, distraction and communication problems.
• Complaints and reports: Reports of mobbing, harassment or violence.
• Exit interviews: The frequency of workload, manager relationship and wear-and-tear themes among reasons for leaving.

Leading indicators: the warning signs of the problem

These indicators help catch the problem before it grows:

• Employee perception surveys: Workload, control over the work, role clarity, manager support, recognition, psychological safety and work-life balance.
• Use of support services: Total numbers, without personal information, of employee assistance programme or counselling applications.
• Overtime and leave use: Teams that work constant overtime or cannot use their annual leave.
• Participation in training and programmes: Which units are being reached and which are not.

The ILO's 2026 report proposes a similar set: absence, turnover, reports of violence and harassment, use of support services and employee surveys (ILO, 2026). Monitoring these indicators together reduces the risk of any single piece of data being misleading.

Do not forget presenteeism

Absence records do not show employees who come to work but cannot work at full productivity. When the cost of poor mental health to employers in the UK was calculated, the largest item was found to be presenteeism (Deloitte, 2024). Adding a question to your survey such as "In the last month, were there days when you could not fully focus on your work for health reasons?" can partly make this invisible area visible. I have covered the subject in detail in the article on presenteeism.

How Do You Design a Good Employee Survey?

The survey is the heart of wellbeing measurement; but a badly designed survey both produces wrong data and damages trust.

1. Clarify the purpose. A "let's ask about everything" approach leads to long, tiring surveys. Decide in advance which decision you will make with this data.
2. Keep it short. Especially for shift workers, being able to complete the survey within a break increases participation.
3. Use plain language. Prefer everyday language to academic terms. Statements such as "I know clearly what is expected of me" are more understandable than "role ambiguity".
4. Use proven tools. Rather than writing your own questions from scratch, drawing on scales whose validity and reliability have been tested makes the results sounder. The tools in the annex of the ÇSGB guide can be a useful guide in this respect.
5. Repeat the same questions. To see the trend, the core questions need to stay the same at every measurement.
6. Leave an open-ended space. The question "What puts the most strain on your wellbeing in this workplace?" can bring out reasons the numbers cannot explain.
7. Equalise access. A survey sent only by email does not reach employees in the field. Offer options such as paper forms, kiosks or QR codes.

The value of a survey depends less on the questions it asks than on the employee believing that their answer will not harm them.

Confidentiality and anonymity

If employees suspect that their answers can be matched to their identity, they either do not take part or give "safe" answers. For this reason, the organisation needs to set a minimum group size so that the results of small groups are not reported separately, keep demographic questions to a minimum, and announce clearly who will hold the data and where. Having the survey run by an independent external organisation can also increase trust. I have explained this subject in detail in the article on trust and anonymity in employee surveys.

How Are Wellbeing Measurement Results Interpreted?

Collecting data is half the job; the real value comes from correct interpretation.

• Do not get stuck on the average. A result that looks acceptable across the organisation can hide a serious problem on a single shift. Compare with suitable sub-groups that protect anonymity.
• Look at the trend. A single measurement is a photograph; two or more measurements are a film. The direction of change is more meaningful than a one-off score.
• Do not misread unexpected increases. An increase in use of support services or in reports after a programme starts usually shows that trust is rising; it may not mean the problem is growing.
• Do not claim causation. If absence fell in a year, your programme may not be the only reason; economic conditions, the season or the production schedule may also have played a part. Report findings in honest language.
• Read quantitative and qualitative data together. The numbers show "what" is happening; focus groups and open-ended comments show "why".

Suppose in a metal-processing factory the survey shows that perceived manager support on the night shift is markedly low, and it is also seen that near-miss reports are low on the same shift. Read together, these two pieces of data suggest the problem may not be simply "dissatisfaction" but safety risks that cannot be spoken about on the shop floor. The action plan then takes shape accordingly.

Common measurement mistakes

Some mistakes that recur even in experienced organisations reduce the value of data collected with great effort:

• Limiting the survey to satisfaction questions. The question "Are you satisfied with our company?" gives no information about changeable conditions such as workload, role clarity or manager support. Questions should focus on areas where action can be taken.
• Leaving measurement to a moment of crisis. A first measurement taken right after a wave of redundancies, a major workplace accident or a merger may not reflect ordinary conditions. If measurement is to be done in such a period, this context should be stated clearly in the report.
• Changing the questions every year. Adding new questions is natural; but when the core indicators change, the possibility of comparison is lost.
• Leaving managers alone with the results. A manager who sees negative results for their team may become defensive or start trying to find out "who wrote that". The results should be discussed with managers in a non-blaming framework, accompanied by a guide.
• Keeping the data only in HR. If the OHS unit's accident and near-miss data and the survey results are not read together, important connections are missed, particularly in factories.
• Polishing the results. Reporting only positive findings to senior management, although reassuring in the short term, prevents the organisation from seeing its real risks.

The most practical way to avoid these mistakes is to tie the measurement process to a written plan from the start: which data will be collected, by whom and when; with whom, and in what form, the results will be shared; and into which decision mechanism they will feed. This plan preserves the continuity of measurement even when managers change.

From Measurement to Action: Completing the Loop

Not sharing the results of a measurement and then having nothing change is the most important factor that lowers participation in the next survey. For this reason:

• Share a summary of the results with all employees within a reasonable time after the survey closes.
• Set a small number of concrete actions for each unit, each with a named owner.
• Plan the actions together with employee representatives.
• At the next measurement, give feedback along the lines of "you told us this last time, and here is what we did".

Measurement gains meaning when it is designed as part of the programme. I explained the overall design of the programme in the article on how to build a wellbeing programme.

How Do You Get Started in Your Organisation?

1. Needs analysis: First take an inventory of the data you already have: absence and turnover in the HR system, accidents and near misses in the OHS unit, and aggregate health surveillance data from the workplace physician. Identify which data is missing and put in writing which decision you want to make with measurement.

2. Programme: Design a short, evidence-based and repeatable survey; announce the confidentiality rules in advance and plan how field employees will access it. Add a few focus group sessions alongside the survey. For the design and delivery of this process, you can look at the analysis options on the corporate services page.

3. Measurement: Carry out the baseline measurement, share the results and put the actions into practice. Re-measure the same indicators at six and twelve months; report the change by comparing it with the baseline. A comprehensive survey once a year, together with short pulse surveys of a few questions in between, helps you monitor the trend without adding to the burden.

Frequently Asked Questions

How often should wellbeing be measured?

A comprehensive measurement once a year with short pulse surveys in between is usually enough. Long surveys carried out too often create fatigue and lower participation. What matters is that action follows measurement and that the results are shared.

Can employees be assessed on the basis of survey results?

No. Wellbeing surveys should be used to understand conditions at team and organisation level; they should not be used in individual performance appraisal or in decisions about employment. Otherwise trust is lost and the data loses its meaning.

Does it make sense to run a survey in a small business?

Because anonymity is harder to protect in small businesses, sub-group analysis should not be done. A short organisation-wide survey, one-to-one conversations and regular monitoring of basic indicators such as absence give a valuable picture even on a small scale.

Are the results useful if participation is low?

Low participation means the results may not represent all employees; it is also a trust indicator in itself. Understanding why participation stayed low, for example by examining concerns about confidentiality, access problems or earlier surveys that led nowhere, is the first step to improving the next measurement.

• #wellbeing measurement
• #employee survey
• #HR analytics
• #psychosocial risk

Yazar

Eda Erkan Alapala

Klinik Psikolog. İstanbul Kent Üniversitesi Klinik Psikoloji Tezli Yüksek Lisans mezunu; bilişsel davranışçı terapi, çocuk merkezli oyun terapisi ve çocuk-ergen psikoterapisi alanlarında eğitimlerini tamamlamıştır.

Eğitimler