Ask any clinic director what keeps them up at night and turnover will be near the top of the list. The numbers in ABA are sobering: industry surveys routinely report annual Registered Behavior Technician (RBT) turnover north of 50%, with much of it concentrated in the first several months on the job. Every departure is a re-hire, a re-train, a re-credential, and — most painfully — a disruption to the learners on that technician’s caseload, who lose a familiar adult and have to re-pair with someone new.
It’s tempting to file turnover under “that’s just the field” and move on. But a large share of early attrition is not inevitable. It’s the predictable result of how we onboard people into one of the most demanding entry-level roles in healthcare. If you understand why people leave in the first six months, you can change the conditions that drive them out.
The story we tell ourselves about turnover (and why it’s incomplete)
The convenient explanation is money: RBTs leave for a dollar more an hour somewhere else. Compensation absolutely matters, and underpaying technicians will cost you good people. But pay alone doesn’t explain the timing. If it were purely about wages, attrition would be spread evenly across tenure. Instead it clusters early — often in the first 90 to 180 days, before a technician has even hit their stride.
That timing points to something else. New RBTs leave early because the gap between their training and the reality of the floor is too wide, and they’re left to close it alone, in front of real clients, under real pressure. The exit interview says “found another opportunity.” The real story is usually some version of I never felt like I knew what I was doing, and no one had time to show me.
What actually drives early attrition
Across the clinics that struggle most with first-year retention, the same preventable drivers show up again and again.
1. The competence–confidence gap. The 40-hour RBT training and the credentialing exam establish baseline knowledge. They do not produce a person who can fluidly run a discrete trial, recover from a refusal, or stay regulated when a learner escalates. New technicians know the terms — DTT, prompting hierarchies, reinforcement — but knowing is not doing. The first time they have to perform a skill is often with an actual client, and the gap between “I read about this” and “I can do this under pressure” is where confidence quietly collapses.
2. Trial by fire. Caseload pressure means new RBTs are frequently running sessions semi-independently within days or weeks. There’s rarely room for low-stakes practice first. A technician’s earliest reps happen on real learners, where mistakes have real consequences — a missed prompt, a reinforced problem behavior, a session that spirals. That’s a high-anxiety way to learn anything, and it teaches new staff to associate the work with the feeling of being underwater.
3. Thin, inconsistent supervision. BCBAs are stretched. Behavior Skills Training (BST) — instruction, modeling, rehearsal, feedback — is the evidence-based way to build technician competence, but its most predictive component, rehearsal, is also the one most often skipped when a supervisor is juggling a full caseload plus billing plus parent meetings. New RBTs get told what to do and maybe shown once, but rarely get to practice with feedback before they’re expected to deliver. Without that loop, they don’t improve fast enough to feel successful.
4. Emotional load without preparation. ABA work involves challenging behavior: aggression, self-injury, elopement, big emotions from learners and sometimes from caregivers. A technician who hasn’t been prepared for those moments experiences them as threatening rather than manageable. Repeated exposure to situations you feel unequipped for is a fast track to the emotional exhaustion that precedes burnout — and burnout precedes resignation. (We’ve written more on burnout’s causes and warning signs if you want to go deeper there.)
5. No early wins. People stay where they feel they’re getting good at something. When a new RBT’s first months are a blur of confusion and correction, they never get the reinforcing experience of visible competence. The work feels like something they’re failing at rather than something they’re mastering. So when a less stressful job appears, leaving feels like relief, not loss.
Notice the through-line: most of these drivers are about preparation and practice, not personality or pay. That’s good news, because preparation is something a clinic can control.
What actually keeps them
Retention strategies that move the needle tend to attack the competence–confidence gap directly and early.
Structure the first 90 days. Don’t leave onboarding to chance and caseload availability. Map out what a new technician should observe, practice, and demonstrate in weeks 1, 2, 4, and 8. Make competence milestones explicit so both the RBT and their supervisor know what “ready” looks like. (Our guide to the first 90 days as an RBT lays out a useful arc.)
Build in reps before real clients. This is the single highest-leverage change most clinics can make. Before a new RBT runs a procedure with a learner, they should have practiced it — out loud, with their hands, in something close to real conditions — and received feedback. Role-play with a supervisor is the classic vehicle. It works, but it’s expensive in BCBA time and limited by how many scenarios a busy supervisor can stage. The goal is simple: the first time a technician handles an escalation or a refusal shouldn’t be the first time they’ve ever attempted it.
Protect the rehearsal step in supervision. If BST is your model — and it should be — treat rehearsal as non-negotiable, not as the step you cut when you’re behind. Even brief, frequent practice-with-feedback cycles beat occasional long observations. Make feedback specific, behavioral, and immediate. (See our piece on effective supervision feedback.)
Prepare people for the hard moments specifically. Don’t let challenging behavior be a surprise. Walk new technicians through what escalating self-injury, elopement, or caregiver pushback actually look like and what their role is in each — before it happens. A technician who has mentally and physically rehearsed a scenario meets it as a professional with a plan, not as a person in over their head.
Engineer early wins. Sequence responsibilities so new RBTs succeed at something real quickly, and name those wins out loud. Competence is reinforcing. The technician who feels themselves getting better is far more likely to still be there at the one-year mark.
The economics, briefly
If the clinical and human arguments aren’t enough, the math usually is. Replacing an RBT — recruiting, the 40-hour training, credentialing, supervisor onboarding time, and the productivity drag while a new hire ramps — costs thousands of dollars per departure, before you count the clinical cost to learners who lose continuity of care. Investing in better preparation isn’t a soft perk. It’s one of the highest-ROI operational moves a growing ABA organization can make, because it attacks turnover at its most preventable point: the beginning.
How Kipr Helps
Most early RBT turnover comes down to a gap between training and the floor — technicians thrown into high-stakes moments before they’ve had a chance to practice them safely. Kipr is built to close that gap. Our AI-powered simulations let behavior technicians and trainees rehearse realistic client scenarios — an escalating learner, an escape-maintained refusal, a caregiver who’s skeptical of the plan — as many times as they need, with no risk to a real client and without consuming hours of a BCBA’s caseload. Practitioners build fluency and confidence before they’re in the room, so their first real attempt isn’t their first attempt at all. That’s how you turn a nervous new hire into a capable technician who stays.
Kipr is in active development for the ABA community. If you lead a clinic or supervise technicians and want earlier access to simulation-based practice that helps your team build confidence before they build it on real clients, join the Kipr waitlist — we’ll keep you posted as we open early access.