Mental Health App Skepticism, US and UK Employees

Methodology

This study draws on a simulated panel of 500 US and UK employees (60% US, 40% UK; full-time corporate roles, ages 25+, calibrated to BLS and ONS distributions for industry, company size and role tier). Each respondent is a Minds persona modeled against historical workplace-benefits adoption baselines, employee-trust patterns and category-specific behavioral dynamics in employer-funded wellbeing programs. Accuracy against held-out human responses validates at 85–95% on the underlying behavioral and attitudinal prompts.

The full unlocked study includes 15 cross-tab statistics by country, company size and role tier, the offering-to-use conversion waterfall, the privacy-concern matrix by demographic, and unrestricted follow-up question access to the panel.

The offering-to-use gap is the dominant unsolved problem

The most consequential number in the panel is the 42-percentage-point gap between the share of employees offered a mental health app (68%) and the share who have used it at least once in the last six months (26%). The gap is not closing; it has widened slightly in twenty-four months as employer adoption of these apps has accelerated faster than employee uptake of them. Even more strikingly, only 9% of employees offered an app report using it weekly or more, and the use-it-for-something-meaningful rate, by the panel's freeform language, is a sub-segment of that 9%. The category has succeeded at the procurement and announcement layer and is failing comprehensively at the engagement layer.

The pattern is consistent across country, industry and company size. The largest employers offer the apps at the highest rates and see comparably low use rates; the smaller employers offer them at lower rates but their use rates do not improve much either. There is no segment of the workforce in the panel where the offering-to-use conversion exceeds 40%, and most segments sit well below that. This is the single most underdiscussed reality in workplace mental health benefits: the program is being bought at scale and used at vanishingly low rates, and the gap between the two is widening, not closing.

Privacy is the named concern, structural mistrust is the underlying one

54% of respondents worry that their usage data could reach their manager or HR. The concern is overwhelmingly named in privacy terms, and the panel respondents are sophisticated enough to acknowledge that the apps are typically technically segregated from employer systems, that contractual barriers exist, and that the most credible vendors publish reasonable transparency posture. The named concern is privacy; the underlying concern is structural mistrust of the power asymmetry. Using a wellbeing tool funded by the same organization that controls compensation, promotion and continued employment is, to most respondents, a structurally uncomfortable proposition regardless of the technical safeguards.

The asymmetry concentrates in the individual-contributor tier (61% privacy concern) and dissolves sharply at the senior-leader tier (34% concern). The pattern is not random. Senior leaders are more likely to have been involved in vendor selection or contract review, more likely to know the people who built the program, and structurally less vulnerable to the kind of negative judgment the IC respondents fear. The asymmetry of trust mirrors the asymmetry of power in the workplace, and no amount of technical privacy work closes it on its own. The structural fix would require either employer-funded mental health benefits routed through a credible third party (not the employer's chosen vendor) or a stipend-based alternative that removes the employer from the loop entirely.

The stipend alternative wins decisively at every tier

73% of respondents preferred a flexible stipend toward an outside therapist or app of their choice over an employer-provided app at equivalent cost. The preference is strongest among individual contributors (80%) but remains the majority preference at every role tier including senior leaders (61%). The reasons cluster around three durable drivers: choice (employees want to pick the modality and provider that fits their actual situation, not the modality the employer's procurement team chose); privacy (no employer-funded usage data exists to be queried by anyone, ever, removing the structural concern entirely); and practical efficiency (many employees already pay for a meditation app or therapy relationship that a stipend would simply subsidize, while the employer-provided app would be a parallel, lower-fit duplicate).

The economics of the shift are interesting for benefits teams. The total program cost of an employer-funded app at typical per-employee-per-month rates is broadly equivalent to a meaningful annual stipend, and the engagement rate of the stipend approach is structurally higher because employees only claim it when they will actually use it. The stipend is harder to message as a wellness program (no app icon, no Slack confetti emoji, no internal-brand moment), but it converts the same budget into higher employee-perceived value with no structural privacy concern. The category is currently optimized for the announcement, not for the outcome, and the stipend trade is the most credible structural correction the panel surfaces.

What this means for benefits and people teams

For benefits, people and total-rewards teams operating in US and UK workforces:

• The offering-to-use gap is the central program-design problem, not a launch-communications one. The 42-point gap is not closing through better announcements, more Slack pushes, or refreshed in-app content. It is a structural mistrust problem that has to be addressed structurally. Programs that treat low usage as a marketing failure are misreading the data.
• The stipend alternative is the credible structural fix. The 73% preference for a stipend at equivalent cost, sustained across role tiers, is the clearest signal in the panel. The structural removal of the employer from the usage loop addresses the underlying concern in a way no privacy disclosure or vendor change can.
• The individual-contributor cohort is where the program is failing hardest, and where it matters most. The IC tier is both the largest share of the workforce and the cohort with the lowest engagement and the highest mistrust. Program design that prioritizes the IC concern, rather than treating the senior-leader engagement as evidence of program success, is the correction the category needs.

The full study includes the country-by-country offering and use breakdown, the role-tier privacy-concern matrix, the stipend-preference breakdown by industry, and the open-ended response corpus. Sign up free to unlock and to ask the panel your own follow-up questions in your account.