Updated
Google Ads bring the highest-intent ABA families, but the click isn’t exclusive: the same parent contacts several practices in one sitting, and whoever answers first with a real answer usually wins. Run a few tightly themed campaigns, add negatives for job and training searches, use call tracking, and cover nights and weekends, when 48% of family contacts arrive.
Search is the highest-intent channel in ABA and the most expensive. A parent typing "aba therapy near me" or "autism services accepting medicaid" has already decided to act. You are not persuading anyone; you are being present at the moment of action, and every competent competitor in your market is trying to be present too.
Which produces the defining feature of this channel: the click is not exclusive. The same parent will contact several practices in one sitting, frequently in the same ten minutes. The auction decides who gets seen. Something else entirely decides who gets the child.
You are buying a place in a comparison, not a customer. That reframing changes what you optimize. Ad copy and bid strategy determine how many parents reach you. Response speed, coverage and intake quality determine what share of those you convert — and the second set of levers is usually further from optimal than the first.
It also changes how you read your own numbers. A campaign producing plenty of calls and few clients is rarely a targeting problem. It is usually a phone problem.
Keep it boring. A small number of tightly themed campaigns, separated by the thing that changes your economics: location, service line, and payer language.
"aba therapy near me," "autism therapy [city]," "applied behavior analysis [county]." Phrase and exact match. This is the money, and it deserves its own budget rather than sharing with experiments.
"aba therapy that takes medicaid," "[plan name] autism coverage." Lower volume, far higher conversion, and it pre-filters the objection that kills most ABA inquiries. Worth more budget than most accounts give it: in our research, 46% of families reported Medicaid as their payer, so this is the median family rather than a niche.
"autism evaluation near me," "how to get my child tested." High volume, no diagnosis yet. Only run these if you have a real evaluation path or in-house testing — otherwise you are paying to say no.
Cheap, high-converting, and worth owning so a competitor does not sit above your own name. Keep it in a separate campaign or it will flatter every other metric you look at.
One campaign per clinic or service radius when the economics differ by site. Averaged geography hides the location that is quietly losing money.
ABA search is polluted with traffic that will never become a client, and an unmanaged account spends a serious share of budget on it. Build the negative list before you launch and revisit search terms weekly for the first two months.
The recurring offenders: employment and training searches ("rbt certification," "bcba salary," "aba jobs," "rbt training"), academic and definitional searches ("what is applied behavior analysis," "aba vs speech therapy," "aba criticism"), searches for other meanings of the acronym entirely, insurance-only searches with no service intent, and out-of-area cities that share a name with somewhere you serve.
Employment traffic is the big one, and it is bigger than most teams believe — in many practices a large share of what arrives through the website is not a family at all, but an RBT or behavior tech looking for work. A practice that is also hiring should run recruitment as a separate campaign with its own budget and its own queue, not let applicants eat a clinical acquisition budget and clinical intake hours.
Match the promise in the ad, exactly. A click on "ABA therapy in [county] accepting Medicaid" should land on a page that names the county and the plan in the first screen, not a generic homepage that makes the parent hunt.
Then remove every step between interest and contact. A tracked phone number at the top, tappable on mobile, where most of this traffic lives. An intake that begins on the page itself rather than a "contact us" form that promises a callback. Clear statements of service area, ages served, and payers accepted, because those three facts are what the parent is comparing across their open tabs. No PDF packet download. No portal registration.
One more thing most practices miss: say what happens next and when. "Someone will call you within the hour, including evenings" converts better than any testimonial on the page.
Most ABA search conversions are phone calls, which means the important half of your performance data lives outside the ad platform by default. Without call tracking you cannot tell which keywords produce calls that become clients, and you certainly cannot tell how many of those calls rang out.
Dynamic numbers per campaign or per source, recordings and transcripts where consent rules allow, and one number that matters more than the rest: the share of paid calls that were actually answered. In most first audits, that number is the finding.
Ads serve around the clock, and families search when they finally have a free moment — after dinner, after bedtime, on Sunday. In our published cohort of 120 providers and 185,000 conversations, 48% of family contacts arrived outside business hours.
There are two honest responses. Restrict ad scheduling to the hours you can genuinely answer, and accept that you have conceded the busiest window in the category to competitors. Or cover the hours the ads run. Dayparting is the cheaper fix and the smaller business. What does not work is running 24/7 ads into a voicemail box and calling the leads low quality.
The first conversation has four jobs: confirm the service area, confirm the payer, confirm the child’s age, and find out whether there is a diagnosis yet. Done well, it takes ninety seconds and feels like help. Done badly, it feels like a screening call and the parent hangs up and calls the next tab.
Lead with what you can do. "We serve that county and we do take that plan — let me get the first piece of this started for you while we talk" is qualification. "What insurance do you have?" as the opening line is an obstacle course.
Three numbers per campaign: cost per started client, share of calls answered, and median time to first human contact. Conversion tracking should fire on outcomes further down the funnel than a form submission — import the later events so smart bidding optimizes toward families who actually start.
Expect lag and plan for it. ABA acquisition-to-start commonly runs weeks, so judging a campaign on a 14-day window will make you cancel the thing that works.
Clicks in this category are expensive because intent is high and several practices compete for the same local searches; costs vary widely by market and by how many providers bid locally. Rather than benchmarking cost per click, track cost per started client — a market with costly clicks and fast response can be cheaper per admitted child than a market with cheap clicks and slow follow-up.
Start with employment and training terms (rbt certification, bcba salary, aba jobs, rbt training), academic and definitional terms (what is aba, aba vs speech therapy), unrelated meanings of the acronym, and out-of-area place names. Employment traffic is usually the single largest source of wasted spend for clinics that are also hiring.
Only if someone or something can answer 24/7. Roughly half of family contacts arrive outside business hours, so around-the-clock ads into a voicemail box pay to send families to competitors. If you cannot cover those hours, restrict ad scheduling to the hours you can answer.
Yes. Most conversions in this category are phone calls, so without call tracking you cannot attribute clients to keywords or campaigns, and you cannot see your unanswered-call rate — which is frequently the biggest problem in the account.
Only if you can do something with those families — in-house testing, or a real referral path to diagnosticians with known waits and payers. Otherwise you are paying premium prices to tell parents to come back in nine months, and most of them will not.
On this page, the auction decides who gets seen and something else decides who gets the child: who answers first, with a real answer. Carelu is that answer, on the first ring and in the first ninety seconds, at every hour the ads are running.
Campaign structure, negatives, landing pages and bids stay with you and your agency. Carelu covers the other half of the account: the calls and forms those clicks produce, and the numbers that show which of them become clients.
An AI voice agent picks up every paid call, 24/7, in English and Spanish. The 48% of contacts that arrive after hours don’t reach voicemail.
Service area, payer, age and diagnosis status are resolved in the first conversation, leading with what you can do. Families without a diagnosis are routed to an evaluation path.
Web forms and chats are answered within seconds, and intake starts during that first contact, with insurance eligibility and ABA benefits checked automatically.
Carelu separates applicants from families on your website. 40% of captured website leads are job seekers, and they shouldn’t use up clinical intake hours.
With Google Ads and CallRail connected, reporting shows answer rate, speed to contact and cost per started client by channel and source.
Two practices in the same market can run near-identical campaigns and admit very different numbers of children. The difference shows up after the click.
You already paid for the click. Answer the call.
Carelu picks up every call, chat and form instantly, qualifies the family, starts intake on the spot, and tells you what each keyword is really costing you per started client.
Get a Demo