Whole can read your calendar, if you connect it. Not the contents of your meetings. The shape of them.
Specifically it derives three things: meeting time falling outside ordinary working hours, meeting time at the weekend, and one-to-ones. Those feed the recovery and support parts of your plan, so a week with six evening calls in it does not get treated the same as a quiet one.
People ask what that is for, and the honest answer needs separating into three parts: what the product observes, what the research supports, and what we merely think.
What Whole actually observes
Calendar connection is opt-in, per source, and revocable without losing the rest of the app. When it is on, Whole classifies meeting time by local wall-clock time. Anything outside roughly eight to six on a weekday counts as after hours. Anything at the weekend counts as weekend work.
That is the whole mechanism. There is no natural-language reading of invitations, no attendee analysis, no productivity inference. Meeting minutes and when they fell.
For an employer, none of it surfaces individually. Calendar-derived signals feed the same aggregate reporting as everything else, above the same participation floor, and no cohort of fewer than 5 people is reported on at all.
What the research actually supports
This is where it is worth being careful, because the obvious story is more confident than the evidence.
A 2024 fixed-effects study of 1,465 Norwegian full-time employees, tracked across four waves over two years, examined four kinds of after-hours work: long daily hours, late evening work, quick returns between shifts, and long weekly hours. Its finding was that long weekly working hours were significantly associated with burnout, work-home conflict and reduced psychological detachment, after controlling for the other three.
The part worth sitting with: late evening work did not independently predict burnout once the other work-hour types were accounted for. Neither did long individual days. The signal was in the weekly total.
Source: Bernstrom VH, Ingelsrud M, Nilsen W. The consequences of after-hours work: a fixed-effect study of burnout, pain, detachment and work-home conflict among Norwegian workers. Scandinavian Journal of Work, Environment and Health, 2024;51(1):38-47.
That complicates the tidy version of this argument. "Evening meetings are a warning sign" is exactly the claim the study does not support. What it supports is narrower and less quotable: sustained weekly volume is what shows up, and a single late night is mostly noise.
What we think, clearly labelled as what we think
We built calendar signals because we believe the shape of someone's week is a legitimate input to their wellbeing plan, and because it is one of the few inputs a person does not have to remember to log.
We do not claim, and cannot show, that calendar load predicts burnout before a survey would catch it. That is a stronger claim than we have evidence for. We considered publishing this piece under a title that said exactly that, and did not, because the research we could find does not support it and we are not going to assert a temporal claim because it makes a better headline.
What we would say is narrower. Weekly load is measurable continuously and cheaply, where a survey is periodic and expensive. If the thing associated with burnout is sustained weekly volume, then something that watches weekly volume is watching a relevant quantity. Whether it is an early indicator is an open question, and one we would need longitudinal data of our own to answer honestly.
Why any of this belongs in a wellbeing conversation
Most wellbeing provision is aimed at what a person does with the time around their work. Sleep better, move more, breathe here, be more resilient.
Work rhythm is the other half, and it is the half an organisation can actually change. An employee cannot unilaterally reduce their meeting load. Someone with authority over how a team's week is structured can.
That is why we would rather an aggregate view showed a People team that recovery signals across a function have been drifting for six weeks, than have the platform spend its energy coaching individuals to cope with something structural. Coping strategies for a workload problem are a way of making the person responsible for the system they are in.
What this does not entitle anyone to conclude
A calendar is not a diagnosis. Meeting load varies for reasons that have nothing to do with strain: a launch, a hiring round, a fortnight of interviews, someone who simply likes talking to people.
Aggregate calendar signals are a reason to ask a better question. They are not evidence about any individual, they are not available about any individual, and they should not be the basis of a conversation that begins with somebody's name.