Drift: catching burnout while it is still a drift
01The problem
Burnout does not happen on a specific day. The World Health Organization describes it as a syndrome resulting from chronic workplace stress that has not been successfully managed, with three dimensions: feelings of energy depletion or exhaustion, increased mental distance from one's job or feelings of cynicism toward it, and reduced professional efficacy (World Health Organization, 2019). Chronic is the important part, since burnout builds up over weeks and months, so by the time someone says they are burnt out, it has usually been building for a long time.
The brief points out that early detection might favour early treatment, which made sense to me, but when I looked at how burnout is usually measured, I noticed that most tools are not built for early detection at all.
02The point I chose
The most widely used burnout questionnaire, the Maslach Burnout Inventory, and most tools based on it, are taken once and compared to a fixed cutoff (Maslach et al., 2001). That works for research, and for checking how bad things are right now, but it has a blind spot. Early burnout rarely crosses a cutoff. It shows up as a slow slide away from what is normal for that specific person, and since each week only looks a little worse than the one before it, neither the person nor a one-time test notices.
So instead of covering everything about burnout, I went deep on one point, which is comparing a person to their own normal week after week, instead of comparing them to everyone else once.
03What Drift is
Drift is a small web tool where you check in once a week, which takes about a minute. It asks six questions, plus one about what weighed on you most that week. After three check-ins it knows what a normal week looks like for you, and from then on it compares every week to that, raising a flag only when a signal stays above your normal range for more than one week in a row. When it does, it suggests what to do based on which signal is moving and what has been weighing on you.
The tool opens on an example, Sara, a made-up employee with twelve weeks of check-ins, so anyone can see what it does before trying it on themselves.
04How it works
- Check inSix questions, answered in workdays from 0 to 5.
- ScoreEach signal becomes a number from 0 to 100.
- BaselineThe first three weeks set your normal range.
- CompareEvery later week is checked against that range.
- ActA status, and advice when something is drifting.
The questions
I wrote two questions for each of the WHO's three dimensions. Each one asks for a number of workdays instead of how much you agree with a statement, since counting days is more concrete and less affected by whatever mood you happen to be in while answering. The Maslach Burnout Inventory is a licensed instrument, so I did not reuse its items, but the three dimensions are the same.
| Question (on how many workdays did you...) | Signal | Scoring |
|---|
Scoring
Each answer becomes a score from 0 to 100, where a higher number always means more of a warning sign. That means the two questions about progress are reversed, so five days of feeling your work made a difference counts as 0. Each signal is the average of its two questions.
The baseline and the rules
The first three check-ins become the baseline. For each signal, Drift takes the average of those weeks and how much they vary, and builds a normal range around the average that is at least 10 points wide on each side, so someone with very consistent weeks does not get flagged for small changes. After that, the status follows these rules:
| Status | When |
|---|---|
| Building baseline | Fewer than three check-ins. |
| Steady | No signal has been above its normal range for two weeks in a row. |
| Worth watching | One signal has been above its normal range for two weeks in a row. |
| Drifting | One signal has been above it for three weeks or more, or two signals for two weeks each. |
One bad week never changes the status on its own. That was a deliberate choice, since a tool that raises an alarm every time someone has a rough week gets ignored within a month, and then it misses the drift it was built for.
What weighed on you
The last question is based on the six areas of worklife, which Maslach and Leiter (2016) describe as the main sources of burnout in the workplace: workload, control, reward, community, fairness and values. Each area appears as a plain choice, like "too much to do" for workload or "effort not noticed" for reward. During a drift, Drift counts which one was picked most often and uses it to choose the advice.
The advice
When something is drifting, Drift shows one action for the signal that is moving most, one for the pressure that came up most often, and from three weeks on, a suggestion to talk to someone, whether that is a person they trust, their manager or a doctor.
05The example, week by week
The table below is generated by the tool itself from Sara's check-ins, so the statuses are the ones Drift would actually show at the end of each week.
| Week | Exhaustion | Cynicism | Reduced efficacy | One-time test | Drift |
|---|
06Privacy and team use
Answers about burnout are personal, so Drift has no server side and no accounts. Every check-in stays in the browser of the person who made it and is never sent anywhere, so not even I can see it, and there is a button to delete everything at any time.
Drift is built for individuals, but it could work for a team without exposing anyone. A manager could ask the team to check in privately each week and share only their status, steady, worth watching or drifting, instead of their answers. That gives the manager the early warning the brief talks about, while the details stay with the person.
07Limitations
The questions and the thresholds are my own, and they have not been validated, so Drift is an awareness tool and not a diagnosis. It also depends on honest self-report, like any questionnaire.
The biggest limitation is the baseline itself. If someone is already burnt out when they start, their first three weeks are already high, and Drift would treat that as normal. To soften this, the one-time cutoff is still shown next to the baseline comparison, and Drift points it out when someone's baseline is already high.
Since check-ins live in one browser, switching devices or clearing site data starts over. Drift also compares check-ins, not calendar weeks, so a missed week is skipped rather than counted, and it needs three weeks before it can say anything useful.
08Reflection
The part that surprised me most was how easy it is to build a tool that alarms people, and how much harder it is to build one that stays quiet until it matters. Most of my decisions ended up being about what not to flag. Burnout is gradual, so a tool for it has to be patient too, and I think that is the main thing I would take from this into managing a team, paying attention to how someone's weeks are changing over time, not just how they seem on a given day.
09References
- Leiter, M. P., & Maslach, C. (1988). The impact of interpersonal environment on burnout and organizational commitment. Journal of Organizational Behavior, 9(4), 297–308.
- Maslach, C., & Leiter, M. P. (2016). Understanding the burnout experience: Recent research and its implications for psychiatry. World Psychiatry, 15(2), 103–111. https://doi.org/10.1002/wps.20311
- Maslach, C., Schaufeli, W. B., & Leiter, M. P. (2001). Job burnout. Annual Review of Psychology, 52, 397–422. https://doi.org/10.1146/annurev.psych.52.1.397
- World Health Organization. (2019, May 28). Burn-out an "occupational phenomenon": International Classification of Diseases. https://www.who.int/news-room/detail/28-05-2019-burn-out-an-occupational-phenomenon-international-classification-of-diseases