AI vs Human: What Actually Converts Admissions in Behavioral Health?

Why missed calls remain one of behavioral health's biggest blind spots00:00

Drew's journey from recovery to leading growth at Rollover Rep03:00

Gary Garth: Hi everybody. Welcome back to The elev8.io Podcast, the thought leadership platform designed for behavioral health executives navigating growth, innovation, and impact. Today we're diving into a topic that affects every treatment center in the country, yet surprisingly few organizations know how to truly measure it. Most providers spend significant time and money generating admissions inquiries — but what happens after the phone rings? How many opportunities are being missed? How many families never receive a response? And how much revenue and patient impact is being lost because of admissions inefficiencies? At the same time, AI voice technology is rapidly entering the conversation, creating both excitement and uncertainty. So the question becomes: is AI the future of admissions, or are we overlooking the value of human connection during the most important conversations a family will ever have? Joining us today is Drew Powers, Chief Growth Officer at Rollover Rep. Drew and his team answer more than 400,000 behavioral health calls annually and help treatment providers improve admissions performance, reduce missed call opportunities, and better connect prospective patients with the care they deserve. Drew, welcome to the show.

Drew Powers: Hey Gary. Good to be here with you.

Why achieving a 100% answer rate is difficult for most facilities07:00

Gary Garth: Why don't you give our listeners a quick background on who you are and how you came to this role?

What the data reveals about missed calls during staffed business hours11:00

Drew Powers: I got sober about 15 years ago, became a drug and alcohol counselor at the facility I got sober at — like a lot of people. Since then I've been an interventionist, worked in sales and business development, and found behavioral health marketing really interesting. I wound up at a place called Sandstone Care as CGO for eight years. We grew from a very small IOP to a multi-state teen and young adult treatment organization. I was overseeing marketing, digital marketing, admissions, and business development. Most recently I joined Rollover Rep, which was actually a longtime customer of mine. I met the founder Stefan many years ago when he was at Recovery Brands. He'd been beating the drum about missed calls for 10 years, and it finally feels like everybody's talking about it — but they're talking about it through the lens of AI. That created an interesting strategic question for a business that's always been human-centered: where does AI come in? Stefan and I were working through that together, and eventually I said, this is something I'd really like to work on full-time.

Gary Garth: At elev8.io we've been a long-term partner with Rollover Rep. We even have internal systems where when a client hits a certain missed call threshold visible in our dashboards, our account managers make an introduction. It's a difficult metric to manage — for both smaller and larger facilities, and depending on budget and channel mix, it's hard to predict. What are the most common challenges you see across your portfolio, and what are the misconceptions about managing call overflow with a high-performing admissions team?

The financial impact of missed admissions opportunities17:00

Drew Powers: For any size team, it's just kind of impossible — or at least highly inefficient — to staff to a 100% answer rate. If you generate 30,000 calls a month at 95% answer rate, that last 5% across 24 hours a day, 365 days a year is going to cost a lot to staff. Even if your number is 300 calls a month and you have one or two admissions people, that last bit could require two more FTEs — doubling your staffing cost just to catch the tail. And the problem is that everybody then automatically discounts that bucket of missed calls. Whether it's overnight, after 10 PM, early morning — we discount them because we're not staffed to them and our experience tells us they're bad calls. We call back in the morning, nobody answers, we leave voicemails. But what we never test is: what happens when we answer the first time? At a basic human level, that's what we're here to do — reach people when they reach out for help. That moment can be incredibly fleeting. Beyond that, you don't judge a call before it rings at 2 PM, so why are you judging one before it rings at 11 PM? We qualify calls after they've been answered, not before.

Why answering missed calls should be treated as a revenue-generating activity21:00

Gary Garth: You mentioned 400,000 calls annually. What does your data show about which calls are actually qualified? Are after-hours calls converting at a lower rate, or is it more nuanced?

Drew Powers: Two really interesting things here. First — it's in our name, so people assume we're mainly for after-hours. But most people, everyone I've ever talked to, miss most of their calls during their staffed hours. Your staffed hours are when most calls come in. Overnight, if you have nobody, the miss rate is 100% — but the volume is small. During the day by volume is where people miss the most. We always encourage operators to think 24-hour coverage, not just nights. Our technology allows us to split scripting based on time of day — warm transfer back to the admissions department quickly during business hours, handle differently overnight. The second thing: there is some correlation with overnight being lower quality, but it's skewed by lower numbers. The acuity of someone calling at 3 AM is often higher than someone calling at 6 PM. And your qualification rate changes not because the calls change — it changes when your service offering changes. The day you open a second location or a detox, more of those calls qualify. So I don't like benchmarking qualification rate too heavily because it's very specific to each provider.

Admissions optimization and the role of outbound follow-up28:00

Gary Garth: When I ask facilities whether they can filter missed calls by source, that usually reveals immediately whether they really understand this. A paid search missed call — that person is not loyal to your brand. They searched addiction treatment, clicked your ad, you didn't answer, and they're on to the next. Industry data shows conversion rate drops by 400% in a matter of minutes with a call-back versus live answer. With someone who's picked up the phone in a moment of genuine intent, not answering is almost marketing suicide. Are most of your prospective clients thinking of this as conversion rate optimization, or as a cost center?

AI vs human interaction in behavioral health admissions32:00

Drew Powers: Answering a missed call is a revenue-generating effort — that principle exists in every business. That's why facilities have an admissions team to begin with. We're just applying the same principle to overflow. Whether outsourced or insourced, answering the missed call is the primary objective. We give money-back guarantees — if you don't see an additional incremental admit in the first month that pays for the service, we'll pay you back. And when your current system is a voicemail, only about 1 in 20 people who miss you will leave a voicemail. So if you miss 20 calls, you have 19 you don't even know about. You're shooting completely in the dark on who to call back first. If one of those is something urgent — a current client in crisis — you have no way to prioritize it.

Gary Garth: Let's talk AI. Everyone and their mother is selling AI right now. We saw some clients jump onto AI voice and then shift back to human-answered calls because it just didn't work. How does your AI voice offering compare, and where do you recommend it?

Why AI cannot replace trust, empathy, and human connection38:00

Drew Powers: Everybody has the same underlying technology. They all run off large language models, they're all fed through Twilio, it's transcribing words and responding based on a knowledge base and prompting. The only proprietary piece is the knowledge base and prompting. Is Rollover Rep better positioned to do that prompting because we've answered millions of calls? Sure, maybe. But that's actually not why people hang up on AI. They hang up because it's AI. They know it's a robot. We all do it — call Comcast and the first thing you're doing is yelling "Agent. Agent." Because we're conditioned not to trust it. And even if AI somehow gets good enough to do everything perfectly — pre-screens, schedules the admission, handles every scenario — I'll still bet my salary for the next five years that you get lower conversion rates than with humans. Because we don't sell a product in this space. We're selling trust, relationship, compassion, hope. Those are human emotions. AI doesn't have them. If you're talking to AI, you already know one thing: it cannot care about you. It can't have any concern about what happens to you. And this is a healthcare industry where people are reaching out in crisis. About 1% of our calls have to be referred to 988 — people calling a treatment center who are contemplating ending their life. 988 doesn't staff their lines with voice AI for exactly that reason.

Hybrid AI + human admissions strategies43:00

Gary Garth: So where does AI voice fit?

Drew Powers: I call it voicemail 2.0. If your alternative was that the person leaves a voicemail, voice AI is dramatically better — it captures structured data, integrates into your CRM, gives you prioritization of who to call back and what happened overnight, can transfer to your agents. It's voicemail on steroids. But is it better than a human doing those same things? No — because a meaningful percentage of people simply will not interact with voice AI. They'll hang up, just like people don't leave voicemails. Where AI genuinely helps is agent-assist: AI listening to the call in real time, coaching the human rep — here's the obstacle, here's the best practice to overcome it, here's the right moment to present value, here's the right time to wrap up. That allows you to hire to a lower threshold and get better conversion rates. You get efficiency without removing the human element. That's where this goes. Not AI replacing admissions teams — AI making admissions teams better.

Re-engaging past inquiries and improving conversion rates47:00

Gary Garth: Fantastic. Drew, last question — if you had a loudspeaker for the whole industry and could send one message to admissions directors and behavioral health executives about improving census right now, what would it be?

Drew's advice for behavioral health leaders navigating market challenges53:00

Drew Powers: A lot of people are struggling right now. It's a difficult time in this industry — more competition, saturated markets, tighter belts. My recommendation: don't cut the things that generate revenue. Businesses are durable, but you have to protect the revenue-generating systems. Find ways to be more efficient. Outsource where it makes sense. Optimize your systems. Make sure you're working with good digital marketing partners, thinking ahead on SEO and the LLM landscape, and betting on reputable directories. But if you're struggling, now is the time to optimize the systems that generate admissions — not cut them.

Gary Garth: Amen. Very well said. Double down on what's driving revenue. Drew, thank you so much for your time and for sharing your wealth of knowledge. I highly encourage everyone listening to take Drew up on the free audit offer — it may shine light on opportunities you didn't know you were missing. Take care, everyone.