What If We’ve Been Scaling ABA the Wrong Way?

The BCBA workforce has grown dramatically. Between 2020 and 2025, the number of certified BCBAs increased by approximately 85%, from 44,025 to 81,566.[1]

Yet demand has continued to grow alongside the workforce. In 2025 alone, 132,307 U.S. job postings required or preferred BCBA/BCBA-D certification—the highest level recorded across 16 years of employment-demand data.[2]

We are producing more behavior analysts than ever. And still, the system is asking for more.

My background in Organizational Behavior Management has trained me to look at problems a little differently. In OBM, when an outcome is not improving as expected, we do not automatically assume the answer is simply to add more people or more effort. We step back, place the relevant variables side by side, and ask which parts of the system may actually be influencing the result.

So when I look at ABA, I see two realities sitting next to each other:

The workforce is growing.

And demand for that workforce is still growing.

That raises a bigger question: Can we solve the ABA access problem simply by continuing to expand the workforce, or do we also need to examine the system that determines how that workforce is used?

What if scaling ABA doesn’t require replicating the same service model thousands of times?

What if it requires rethinking who implements intervention, where intervention happens, and what the behavior analyst’s role should be?

Perhaps we don’t only have a workforce shortage.

We may also have a delivery-model problem.

The Scaling Equation

During my first year as a master’s student, I became curious about more than the clinical side of ABA. I also wanted to understand the business behind it: How do ABA organizations grow? What drives capacity? And, ultimately, what determines how many families we can serve?

As I learned more, I began to recognize a relatively straightforward equation behind much of the field’s growth:

More families → more direct service hours → more technicians → more supervisors → more infrastructure.

There is nothing inherently wrong with this equation. It has helped many individuals and families access evidence-based behavioral intervention, and direct intervention will continue to be necessary and appropriate for many people.

But I also began to see its structural limitation: when serving more families depends primarily on adding more clinical hours and more people to deliver those hours, access will always be constrained by the availability of that workforce.

And clinical labor is finite not only in the United States, but around the world.

Organizations can recruit and train more technicians. Universities can prepare more Behavior Analysts. Providers can open additional clinics. Technology can improve scheduling and documentation.

All of those things matter. But they are primarily attempts to make the existing system bigger.

What if we also asked whether the system itself could work differently?

What If We Scaled Expertise Instead of Hours?

One of the most interesting opportunities in behavioral healthcare may be shifting the unit we are trying to scale.

Instead of asking:

How many hours of intervention can a clinician or technician deliver?

We might also ask:

How effectively can behavioral expertise be transferred to the people already present in an individual’s everyday life?

Caregivers are there during breakfast.

During bedtime.

During transitions.

At the grocery store.

During play.

When communication breaks down.

When a child needs to learn a new daily living skill.

These moments happen whether a therapist is present or not.

Research gives us reason to take this opportunity seriously. A 2023 meta-analysis of randomized controlled trials of parent-implemented interventions for autistic children found moderately strong overall benefits, including improvements across social behavior, communication, challenging behavior, and, to a lesser extent, adaptive skills.[3]

This question is also deeply personal to me. I am both a behavior analyst and the mother of two autistic children. I have experienced firsthand what happens when clinical knowledge does not end with a therapy session, but becomes something a caregiver can understand, practice, and use—with appropriate coaching and support—within everyday family life.

That experience changed the way I think about where intervention happens and, perhaps more importantly, where learning opportunities happen.

Caregiver-mediated and caregiver-led approaches create the possibility of moving intervention beyond the boundaries of the clinic or scheduled therapy session and into the environments where behavior naturally occurs.

But this does not mean turning parents into therapists.

And it certainly does not mean giving families a treatment plan and expecting them to figure it out.

It means reconsidering what clinical expertise looks like when the behavior analyst is not always the person directly implementing the intervention.

The clinician’s role begins to shift: from being primarily responsible for direct implementation to also developing the competency of the people who support the individual throughout everyday life.

In that model, the clinician increasingly becomes a designer, educator, coach, analyst, and implementation partner.

Caregiver-Led Does Not Mean Clinician-Less

Reducing dependence on continuous professional implementation should never mean reducing clinical oversight. In fact, effective caregiver-led care may require behavior analysts to develop more sophisticated skills in some areas. They must assess not only the learner, but also the environment in which an intervention will actually be implemented.

They must teach effectively.

Observe caregiver implementation.

Provide meaningful feedback.

Identify barriers.

Adapt strategies to family routines and cultural contexts.

Monitor competency.

And recognize when something is not working.

This is where my previous career in Learning & Development and Organizational Behavior Management has deeply influenced how I think about behavioral healthcare.

For more than two decades, I worked on learning and organizational initiatives for global companies. One lesson from that world has stayed with me:

Providing information does not guarantee implementation.

Someone can attend training and still be unable to perform the skill. They can explain a procedure and still struggle to apply it. They can demonstrate competency in a structured environment and fail to generalize it when the environment changes. The same challenge exists when we work with caregivers.

We cannot assume that because we explained an intervention, modeled it once, or reviewed it during a parent-training session, it will suddenly become part of everyday family life.

That gap between knowing and doing is where implementation science becomes particularly relevant.

Training Is Not the Same as Implementation

Implementation science asks a deceptively simple question:

How do we get evidence-based practices to actually work in real-world settings?

That question matters enormously when caregivers become active intervention partners.

Families differ—their routines, stressors, cultures, learning histories, confidence, resources, and the amount of coaching they may need.

If we want caregiver-led models to scale responsibly, we need to move beyond simply asking whether parent training occurred.

We should be asking:

Can the caregiver demonstrate the skill?

Can they use it without continuous prompting from the clinician?

Can they apply it during natural routines?

What barriers interfere with implementation?

Does the strategy fit the family’s environment?

Does competency maintain over time?

And perhaps one of the questions I find most important:

Do we understand how this caregiver learns before we decide how to teach them?

That question contributed to my development of the Caregiver Learning Evaluation and Readiness Assessment (CLERA™), a framework designed to help clinicians understand caregiver learning readiness, strengths, barriers, and teaching needs before and throughout coaching.

But assessing how a caregiver learns is only one part of the equation.

Behavior analysts routinely select assessments, curricula, goals, and intervention strategies according to the needs of the individual they serve. Caregiver-led care requires the same clinical intentionality.

We should not only individualize how we teach the caregiver. We should intentionally select what we teach and which evidence-based caregiver intervention or framework best fits the family’s clinical goals.

The field already offers approaches designed for different purposes.

RUBI, for example, is a structured parent-training program developed to address disruptive behavior in autistic children. In a large randomized clinical trial, structured parent training produced greater reductions in disruptive behavior than parent education alone.[4]

Project ImPACT is a parent-mediated intervention focused on social communication, with research demonstrating improvements in parents’ use of intervention strategies and associations with children’s spontaneous communication.[5]

BCEM™, the model I developed, takes a broader caregiver-competency approach, teaching caregivers to apply behavioral principles and evidence-based strategies within everyday routines.

These approaches should not be viewed as interchangeable versions of “parent training.” They can represent different tools for different clinical needs.

Just as we would not select the same curriculum, goals, or intervention procedures for every client, we should not assume that every caregiver or family needs the same caregiver-led intervention.

A more intentional pathway might look like this:

Assess the learner → identify clinical goals → assess caregiver learning and implementation needs → select an appropriate caregiver-focused intervention or framework → individualize coaching → measure competency and outcomes → adapt.

That is fundamentally different from simply providing parent training.

Telehealth Is Infrastructure, Not the Intervention

Telehealth creates another opportunity and another common misunderstanding.

It can reduce geographic and transportation barriers, connect specialists with families at a distance, and allow clinicians to observe implementation within natural environments.

But putting a clinician and caregiver on Zoom does not, by itself, create a scalable model of care.

Telehealth is infrastructure. It is not the intervention.

Nor does virtual caregiver-led care mean a caregiver, and sometimes a child, must sit in front of a screen with a clinician for hours.

The screen is one way for the clinician to assess, teach, observe, coach, and provide feedback. Much of the implementation can happen between sessions, within naturally occurring routines.

This isn’t merely hypothetical. ABA research has demonstrated that parents can conduct behavioral assessment and intervention procedures in the home with clinician coaching delivered through telehealth.[6]

The distinction is important:

Caregiver-led describes who is being empowered to implement intervention. Telehealth describes how professional support is delivered. They are not the same thing.

Caregiver-led services can be delivered virtually, through hybrid care, or entirely in person. The appropriate modality should reflect clinical needs and family characteristics, while also considering geography, resources, payer requirements, and the service architecture an organization can provide.

Regardless of modality, the clinical cycle remains:

Assessment → individualized caregiver learning → clinician coaching → caregiver practice → observation → feedback → data → adaptation.

Technology can reduce many geographic barriers.

Implementation is harder.

A protocol that works in one environment cannot simply be translated, placed on a screen, and assumed to work somewhere else. Scaling evidence-based intervention requires understanding the people who will implement it, the environments in which it will live, and the systems built to support it.

But What About Quality?

Whenever we discuss scale in healthcare, there is a question we should ask immediately:

What happens to quality?

Scaling ABA cannot mean lowering clinical standards so that we can reach more people. Nor should caregiver-led care automatically become a lower-cost substitute whenever direct services are unavailable.

The better opportunity is to determine which service model is appropriate for which family, at what intensity, and at what point in care.

That may require expanding what we measure.

Alongside learner outcomes and treatment fidelity, caregiver-led models may need to systematically examine caregiver competency, implementation accuracy, independence, generalization, maintenance, learning readiness, and the quality of clinician coaching.

Because if the mechanism through which intervention reaches the learner changes, our quality systems may need to change with it.

Lower intensity cannot become synonymous with lower quality.

Scale without measurement is risky. But scale with clinical oversight, competency-based coaching, implementation data, and clearly defined decision rules may allow us to expand access without assuming that every family requires the same service-delivery architecture.

Maybe the Future Isn’t One ABA Model

This conversation should not become direct ABA versus caregiver-led ABA.

The future may instead involve a continuum:

High-intensity direct ABA ← Hybrid care ← Caregiver-led ABA ← Consultation and coaching

Different individuals may require different levels of support, and those needs can change over time.

The more useful question is:

Who needs what level of support, delivered by whom, at what intensity, for how long—and what data tell us when that should change?

Using the Workforce Differently

This brings us back to scale.

Caregiver-led care does not eliminate the need for high-intensity ABA. Some individuals will continue to need and benefit from substantial direct implementation.

But not every individual or family necessarily needs the same intensity, delivery model, or number of professional service hours.

Recent research adds important nuance to the relationship between treatment intensity and outcomes. In a 2026 study of 725 autistic children receiving ABA, Samelson, Pfingston, and Sneed found that greater treatment dosage was not associated with greater improvement in adaptive behavior, while dangerous behavior decreased over time regardless of dosage. Treatment dosage was associated with goal attainment, and the authors concluded that the question of optimal ABA dosage remains open.[7]

The question of who implements intervention may be equally important.

Earlier community-based research by Sneed, Little, and Akin-Little compared parent-mediated and paraprofessional-mediated ABA. Because families selected their service model, the study should not be interpreted as evidence that one approach is universally superior. But it contributes to an important discussion about whether effective behavioral intervention must always depend on extensive professional implementation.[8]

Together, these questions point toward a different approach:

“What if instead of defaulting to a particular number of hours or a single delivery model, we became better at matching intensity, intervention model, and level of professional implementation to each individual and family?”

Consider what that could mean for the workforce.

Individuals who require high-intensity direct services could continue receiving them, supported by the clinicians and technicians necessary to provide that level of care.

At the same time, when clinically appropriate, other families could receive lower-intensity caregiver-led services. Clinicians could spend more of their time assessing, designing intervention, teaching, coaching, observing implementation, analyzing data, and making clinical decisions, while caregivers apply the skills they develop throughout everyday life.

That changes the capacity equation.

A clinician who must personally deliver or supervise many hours of implementation for each client has a natural limit on the number of families they can support. With appropriately selected families, lower-intensity caregiver-led models may allow clinicians to extend their expertise across more families while preserving higher-intensity workforce capacity for individuals who need it most.

That is not about doing less ABA. It is about using behavioral expertise differently.

The BCBA workforce has grown substantially, and demand has continued to grow alongside it. We can and should continue developing that workforce.

But workforce expansion may be only one part of the solution.

What if part of a behavior analyst’s impact were measured not only by what happens while the clinician is present, but also by what continues to happen when they are not?

What if our most highly trained clinicians could spend more of their time doing the work that specifically requires their expertise assessing, analyzing, designing, teaching, and making clinical decisions, while appropriately trained and supported caregivers develop competency to implement evidence-based strategies throughout everyday life?

Maybe we haven’t been wrong about how to deliver ABA.

Maybe we’ve been too narrow about what it means to scale it.

The next era of ABA may not be defined by how many hours of therapy we can deliver, but by how effectively we match people to the right intensity and delivery model—and extend behavioral expertise into the environments where life actually happens.

If our goal is to make effective behavioral care available to more people, perhaps the question isn’t only:

How do we build a bigger workforce?

Perhaps it is also:

How do we build a system that uses the workforce we have differently?

References