There's a well-cited piece of research by Gloria Mark at UC Irvine, originally from 2008 and replicated several times since: after a workplace interruption, it takes the average knowledge worker twenty-three minutes and fifteen seconds to fully return to the original task. The number is brutally robust. It holds across roles, ages, industries.
The average small-business operator switches between two and four 'apps' (inbox, calendar, books, CRM, social, etc.) every fifteen minutes of focused work. Even if half of those switches are quick check-ins, the daily refocus cost is somewhere between three and five hours. Out of an eight-hour day.
What we measured
We watched twenty-four of our prospective customers work for a full day each, before they signed on. Median number of app switches per hour: nineteen. Median time the operator spent inside any one tool before switching to another: 3 minutes and 41 seconds. Median total time per day spent on the act of switching itself: 2 hours 18 minutes.
Two and a quarter hours a day is the tool tax. It's not theoretical — we timed it.
Why this matters for OS
Most discussions of automation are about hours saved per task. That's the wrong frame. The bigger savings are the hours not lost to switching between tasks. An OS doesn't just do the work faster; it removes the switching cost almost entirely, because the operator stops needing to be the integrator.
If you knock the refocus cost down from 2 hours 18 minutes to 30 minutes — which is roughly what our customers report by week three — that's the same as adding an extra fifteen weeks of productive time to the year. For one person. Not 'time you could in theory spend better'. Time you couldn't spend at all before.