Ask a departing RBT why they're leaving and you'll hear about pay, burnout, or a return to school. Ask their calendar and you'll often get a more specific answer: three weeks of split shifts with unpaid gaps in the middle, a paycheck that didn't match their hours twice this quarter, forty minutes of unpaid note-writing after the last session of the day, and a cancelled session they drove thirty minutes to discover. Turnover in this field gets discussed as a culture problem. A lot of it is an operations problem wearing culture's clothes.
That's actually good news, because operations can be fixed with systems, this quarter, without waiting for the labor market to change.
Schedule fragmentation. An RBT scheduled 9-11 and 2-4 worked four billable hours and lost a whole day. Fragmented schedules are sometimes unavoidable, but often they're just the path of least resistance for whoever builds the calendar by hand. Scheduling that optimizes against staff availability, client availability, and location, and shows the gaps it's creating, turns dead hours back into either billable time or actual time off. Auto-scheduling isn't about saving the scheduler's afternoon; it's about not quietly billing your techs' patience. The scheduling page shows what the calendar checks while you build it.
Pay friction. Nothing corrodes trust like a wrong paycheck, and most wrong paychecks in ABA come from re-keying: sessions in one system, hours in another, payroll in a third. When hours flow from GPS-verified clock-ins straight into a payroll-ready file, overtime computed with your state's rules, travel time captured, the paycheck argument disappears. So does the Thursday-night ritual where your ops manager reconciles three tabs. Self-serve visibility matters too: an RBT who can open "My Hours" on their phone and see exactly what this check will contain doesn't need to wonder.
Documentation overtime. Unpaid or after-hours note-writing is one of the most cited frustrations in the field. Structured data collection helps twice: capture during the session means less to reconstruct after it, and AI-drafted narratives built from the session's actual data turn a twenty-minute write-up into a three-minute review-and-sign. That's an hour a day back for a full-time tech.
Cancellation chaos. Late cancellations are revenue leaks and morale leaks simultaneously. Tracking late-notice patterns tells you which families need a conversation; instant coverage visibility means a cancelled 2pm becomes an offered make-up session instead of an unpaid hole in someone's Tuesday.
Replacing an RBT costs real money: recruiting, onboarding, credentialing, weeks of reduced caseload while a new tech ramps, and the soft cost of a client relationship reset, commonly estimated at several thousand dollars per departure. A practice of thirty techs at typical field turnover rates loses more to replacement costs in a year than the annual price of any software platform on the market. Retention isn't the HR line item; it's the margin line item.
Pick the leak your techs complain about most, anonymous pulse question if you're unsure, and fix that one first with a system, not a memo. Then measure: schedule density per tech, payroll corrections per cycle, average note-completion time, and late-cancellation rate are all trackable, and all move within a quarter when the tooling changes. Culture work matters too. It just works better when the paycheck is right and the Tuesday hole is gone.