Capacity is not a mood or a self-assessment of wellbeing. It is a scored measure across four domains — energy, recovery, focus and resilience — and measuring it well depends less on the questions you ask than on how you design, time and protect them.
Capacity asks whether a person can sustain their performance — not whether they want to. It is scored across four domains, each 0–100, from statement-level responses.
Most capacity measurement fails at the statement, not the analysis. Three design rules matter more than the rest.
Behavioural, not evaluative. “I have woken up feeling rested” is answerable from memory. “My wellbeing is good” asks the respondent to score themselves against a standard they have to invent, which is why such items correlate more with self-image than with any observable outcome.
Bounded in time. Every capacity statement carries an explicit recent window — typically the past two weeks. Without it, respondents silently choose their own reference period and you lose comparability between people and between cycles.
Universal, not localised. The same statement set runs across every role, level and market, so a score in one team means the same thing as a score in another. Bespoke per-department questions feel more relevant and destroy comparability.
The full statement set, response scale and completion flow are described on the employee capacity survey.
Capacity moves faster than engagement. Annual measurement is too slow to catch a decline while it is still reversible, and continuous pulsing produces drift you cannot act on.
A repeating cycle — measure, act, measure again — is what makes the number useful: the second measurement tells you whether the action worked, which a one-off survey never can. The unit of value is the change between cycles, not the score itself.
How measurement, action plans and learning connect across a cycle: how it works.
Capacity data is more sensitive than engagement data — it touches sleep, energy and strain. If people suspect their answers are identifiable, they moderate them, and the measure degrades precisely where it matters most.
Two rules protect it. No result is displayed for a group smaller than five, at any level of the hierarchy, including through cross-filtering. And individual responses are never visible to any manager — only aggregates. These are not policy preferences; they are what makes honest answers rational for the respondent.
Storage, retention and processing detail: trust and security.
A capacity score on its own changes nothing. What makes it actionable is knowing which domain is lowest, which drivers sit beneath it, and who can actually change them — some sit with the manager, some with the organisation, and some with the individual.
Across our validation study (n=10,000+), the separation between the bottom and top decile of the Thrive Index was substantial on outcomes leaders already track: 21% → 98% could regularly perform at their best, and 15% → 98% were not regularly overwhelmed.
Every figure, with its comparison basis: the evidence base.
This page is the method. For the definition, the four domains and the evidence behind them, see employee capacity.
Figures on this page come from thrive at work’s own validation study (n=10,000+, 2025–26), fielded across the US and Europe. Unless stated otherwise, each comparison is bottom decile versus top decile of the Thrive Index.
The domain structure is grounded in peer-reviewed work across physiology, occupational health and organisational psychology; the instrument itself is proprietary.
A 30-minute call with the research team covers what a first cycle would surface in your organisation, and how your sector compares.