Employer wellbeing provision has expanded enormously, and the evaluation evidence is considerably less favourable than the investment implies.
What has been found
Large studies of individual-level wellbeing interventions in workplaces have generally found small or absent effects on the outcomes measured.
Which includes app-based interventions, resilience training, mindfulness sessions and wellbeing days.
One large study examining many organisations found essentially no difference in wellbeing outcomes between participants and non-participants across most interventions offered.
The exception
Volunteering opportunities showed a positive association in the same analysis.
Which is an interesting result and may reflect selection rather than effect.
Why individual interventions underperform
Occupational health research consistently finds larger effects from organisational interventions than from individual ones.
Which follows from the model — if the problem is a mismatch between person and job, changing one side addresses half of it.
The identified drivers of workplace strain include workload, control, reward, community, fairness and values alignment.
Resilience training addresses none of them.
The substitution concern
Offering wellbeing provision can substitute for addressing conditions.
Which is documented in critical analyses of the field, and it locates the problem in the individual rather than in the work.
An employee offered meditation for workload-driven exhaustion receives an implicit message about where the problem lies.
What has better evidence
Job design changes increasing autonomy and control.
Workload management.
Manager training, since manager behaviour is consistently identified as a major determinant of employee experience.
Clear role expectations, since ambiguity is a documented stressor.
And genuine flexibility over when and where work happens.
Why organisations do the other thing
Individual interventions are cheaper, faster, visible and do not require changing anything about how the organisation operates.
Which is a coherent explanation, and it is not a criticism of the people delivering them.
Organisational change requires leadership commitment and produces disruption, which individual programmes avoid entirely.
Measurement
Programmes are frequently evaluated on uptake and satisfaction rather than on outcomes.
Which measures whether people used it and liked it rather than whether anything changed.
Comparing outcomes between participants and non-participants, accounting for who chooses to participate, is what produces usable evidence.
What employees can do
Use what is offered where it is genuinely useful, including employee assistance programmes which are confidential and underused.
And raise workload and conditions through whatever channels exist, since those are where the evidence points.
Anyone experiencing persistent low mood, anxiety or exhaustion should see a doctor rather than relying on workplace provision.
Selection effects
People who take up wellbeing provision differ from those who do not, generally being already more engaged and in better health.
Which means simple comparisons of participants with non-participants overstate effect substantially.
Studies accounting for this find smaller effects, and this explains part of the gap between vendor claims and independent evaluation.
Screening programmes
Health screening offered at work sits in a different category and carries its own considerations.
Which includes whether the screening is one that evidence supports offering to the population concerned.
Screening that is not indicated produces false positives, anxiety and unnecessary investigation, which is a real cost.
Sickness absence and return
Managing return to work after illness is where employer action has clearer evidence.
Which includes phased return, adjusted duties and maintaining contact during absence.
Early intervention consistently outperforms waiting, since the probability of return falls sharply with the length of absence.
Occupational health services support this and are underused, particularly in smaller organisations.
Psychological safety
Whether people can raise problems, admit mistakes and disagree without adverse consequence.
Which is associated with team performance in organisational research and with willingness to report safety concerns.
It is determined largely by manager behaviour and is not addressed by any wellbeing programme.
What to ask of a programme
What outcome it claims to affect, what evidence exists for that claim, how uptake and outcome will be measured, and what happens to the data.
Which are reasonable procurement questions and are frequently not asked.
Confidentiality of any data collected is the one employees care about most and should be established explicitly.
Line manager capability
Manager behaviour appears consistently as the largest single workplace determinant of employee experience.
Which makes manager selection and training higher-leverage than any programme aimed at employees.
Many managers are promoted for technical performance with no training in managing people, and the consequences are predictable.
Small employers
Most people work for organisations too small to run wellbeing programmes at all.
Which means the entire discussion applies to a minority of workplaces.
For small employers the practical levers are the ordinary ones — reasonable workload, clear expectations and flexibility where possible.
Consultation
Asking employees what would help before purchasing a programme is uncommon and cheap.
Which frequently produces answers about workload, staffing and predictability rather than about provision.
Organisations that ask and then buy an app anyway damage trust more than doing nothing would.