Why Can't Therapists Answer Their Phones?
The problem isn't that therapists ignore calls. It's that the work makes answering impossible.
During a 50-minute session, phones ring. Between sessions, there's barely 10 minutes to write notes, use the bathroom, and prepare for the next client. By end of day, most therapists are emotionally spent. The last thing anyone wants is to field intake calls at 6pm from potential clients asking if you take Medicare or treat anxiety.
One Melbourne psychologist described the cycle perfectly: "I hate phone calls. I can't answer during sessions, there isn't time between sessions, and by end of day I'm exhausted. I literally just want someone who can take a call and write down the important info."
The cost of this problem isn't just missed calls. It's lost clients. Research from the Australian Psychological Society shows that 60% of people seeking therapy contact multiple practitioners. The first one to respond usually gets the client. If you're in session for three hours straight, you've already lost those prospects to someone else.
What Are the Traditional Options?
Most therapists try one of three approaches, and all have problems.
Option 1: In-person receptionist. At $25-35 per hour (Australian minimum wage is $23.23), this costs $800-1,400 per month for 20 hours per week. For solo practitioners, this is unsustainable. You'd need to see 10-15 extra clients monthly just to break even. Most therapists can't fill that many additional slots, and if they could, they wouldn't have capacity issues to begin with.
Option 2: Virtual receptionist services. Services like Ruby or Smith.ai charge $325-950 per month for human answering. These work well for law firms with high-value clients, but for therapists billing $150-250 per session, the maths doesn't work. You're paying 2-6 therapy sessions worth of income just to answer phones. They're also US-based, which creates privacy compliance issues under the Privacy Act 1988 when handling Australian health information.
Option 3: Voicemail and callbacks. This is free but terrible for conversion. People in crisis don't leave voicemails. People comparison shopping move on to the next therapist. By the time you call back, they've either found someone else or lost momentum. One Sydney therapist reported that only 30% of voicemails actually convert to booked appointments, compared to 70% of live-answered calls.
How Does AI Handle Therapy-Specific Intake?
AI answering services work differently from virtual receptionists. Instead of a person reading a script, the system uses natural language processing to conduct intake conversations. For therapy practices, this means the AI asks specific screening questions that determine whether someone is a good fit.
Here's what a therapy intake call looks like with AI:
Caller reaches your practice number while you're in session
AI answers immediately (no rings, no hold music)
System asks what brings them to therapy (anxiety, depression, trauma, relationship issues)
AI confirms insurance details or private pay preference
System checks availability preferences (morning, afternoon, evening, specific days)
Caller receives confirmation and you get a detailed intake summary via email or SMS
The entire conversation takes 2-3 minutes. The client feels heard, you get qualified information, and nobody sits on hold.
Most AI answering services only read your website to learn about your practice. This creates gaps. If your site doesn't explicitly list EMDR or mention that you don't take Medicare, the AI can't answer those questions accurately.
Some systems like Trillet use a more comprehensive research framework that automatically scans your website, Psychology Today profile, Google reviews, social media, and even watches video content to understand your practice. When someone calls asking about specific modalities or insurance, the system already knows the answer from multiple sources. This eliminates the awkward "let me check and get back to you" responses that make practices look unprofessional.
What About Crisis Calls?
This is the question every therapist asks, and it's the right one to ask. AI answering services are not crisis intervention tools. They're intake systems. The distinction matters.
When someone calls in immediate distress, the AI should recognise urgency markers in conversation (mentions of self-harm, acute panic, recent trauma) and immediately escalate. Most systems do this in two ways:
Forward to your mobile immediately. If you're available, you take the call. If not, it goes to voicemail and you get an urgent SMS alert.
Provide crisis resources. The system verbally provides Lifeline (13 11 14), Beyond Blue (1300 22 4636), or emergency services (000) while simultaneously alerting you.
You configure these escalation protocols during setup. For example, you might set rules like: forward all calls mentioning "suicide" or "emergency" to your mobile between 9am-5pm, and provide crisis line numbers outside those hours. The system follows these rules exactly, every time.
This isn't perfect, but it's significantly better than voicemail, which provides zero crisis support. And it's better than hiring a receptionist who isn't trained in crisis assessment and might miss warning signs.
Does This Comply with PACFA Ethics and Privacy Act Requirements?
Australian therapists operate under two key compliance frameworks: PACFA ethics standards for counsellors and psychotherapists, and the Privacy Act 1988 for handling health information. Both matter when using AI answering services.
Under PACFA ethics, therapists must maintain confidentiality and informed consent. The question is whether using AI to answer phones violates these principles. The answer depends on what information the AI collects and how it's stored.
Basic intake information (name, contact details, reason for seeking therapy, availability) is generally acceptable. This is equivalent to what a receptionist would collect and doesn't constitute a therapeutic relationship yet. The AI isn't providing therapy or advice, just gathering information to facilitate scheduling.
However, therapists should avoid AI systems that record detailed clinical information during intake. If someone starts describing traumatic experiences or specific symptoms, that crosses into health information that requires stronger protections. Good AI systems recognise this boundary and keep intake questions appropriately surface-level.
For Privacy Act compliance, the key issue is where data is stored and who has access. Australian health information should stay in Australia or with providers who have equivalent privacy protections. US-based answering services create problems here because the CLOUD Act allows US government access to data, even if stored overseas. This violates Australian Privacy Principle 8, which requires reasonable steps to protect personal information from overseas disclosure.
Melbourne-based AI services like Trillet store all data in Australian data centres and comply with local privacy requirements. This matters less for basic intake calls but becomes critical if the system handles appointment notes or session reminders that reference specific conditions.
Victoria has additional requirements under the Health Records Act 2001, and NSW has the Health Records and Information Privacy Act 2002. Both require health service providers to notify clients when collecting health information and explain how it will be used. Practically, this means your AI answering system should have a disclosure statement built into its greeting, something like: "This call will be recorded for intake purposes and shared with [Your Practice Name] to schedule your appointment."
How Does This Integrate with Practice Management Software?
Most therapists use practice management software like Cliniko, Power Diary, or SimplePractice for scheduling and client records. AI answering services need to work with these systems, not create extra administrative work.
Basic integration sends intake information to your email or SMS. You manually enter it into your practice software. This works but creates a 5-10 minute task after each call.
Better integration connects directly to your calendar system. When the AI collects availability preferences, it checks your actual schedule and offers specific time slots. The client can book directly during the call, and the appointment appears in your system automatically. This eliminates phone tag and reduces no-shows because the client chose their own time.
Some therapists worry this removes the human element from booking. In practice, most clients prefer it. They're calling outside business hours because that's when they have time to think about therapy. Being able to book immediately, rather than waiting for a callback, improves conversion. One Brisbane practice reported their booking rate increased from 55% to 78% after implementing direct calendar integration.
Advanced integrations can sync with CRM systems to track where leads come from and measure marketing effectiveness. This is available on higher-tier plans from providers like Trillet (contact required for CRM access), though most solo practitioners don't need this level of tracking.
What Does This Actually Cost?
In 2026, AI answering services range from $29 to $300 per month depending on features and call volume.
Entry-level services like Trillet charge $29 per month with 150 minutes included and no additional telephony fees (the phone system costs are already covered). For comparison, that's less than one therapy session. Most therapy practices receive 15-25 intake calls per month averaging 3-5 minutes each, so 150 minutes covers typical volume. After that, calls continue at no per-minute charge.
Mid-tier options like My AI Front Desk run $249 per month with more customisation and appointment booking features. These make sense for group practices with high call volume and complex scheduling needs.
Human answering services start at $325 per month (Ruby receptionist) and go up to $950+ for full-service virtual reception. These provide a human touch but cost 10-30x more than AI alternatives.
Some marketing agencies and AI resellers charge $1,000+ in setup fees plus $200-500 monthly. These typically configure white-labeled platforms manually, which takes days of work. Trillet's automated multi-source research framework (website, reviews, social media, videos) eliminates this manual configuration entirely. The system learns your practice automatically in 5-10 minutes with no setup fees. Most competitor AI systems only scan your website, so they still require manual setup to fill knowledge gaps.
The return on investment is straightforward. If the system helps you book three additional clients per month at $180 per session, that's $540 in new revenue. Even at $249 per month for a premium service, you're ahead by $291. The real benefit isn't just revenue though, it's reclaiming your time. Not answering phones between sessions means actual breaks. Not returning calls at night means better work-life boundaries. Not wondering if you missed an urgent call means less anxiety about your practice.
What Happens During Setup?
Traditional answering services require lengthy onboarding. You fill out forms describing your practice, list out FAQs, train the receptionist on your policies, and wait days for everything to go live. Most AI systems improve this somewhat by scanning your website, but you still need to manually configure answers to common questions if that information isn't explicitly on your site.
Trillet's approach eliminates most manual setup through its proprietary research framework. You provide your website URL and phone number. The system automatically scrapes your site, reads your Psychology Today profile, checks Google reviews, scans social media posts, and even watches video content to build a comprehensive knowledge base. This multi-source research catches details that website-only systems miss. Setup takes about 5 minutes instead of days.
You then configure three things:
Call routing: When should the AI answer (always, only when you're busy, only after hours)? Most therapists use conditional forwarding where their phone rings first, and AI picks up only if unanswered after 15-20 seconds.
Intake questions: What information do you need from callers? Standard questions include presenting issue, insurance/private pay, preferred availability, and how they found your practice.
Escalation protocols: When should calls forward to you immediately? Set rules for crisis language, existing client calls, or specific keywords that require immediate attention.
After this initial configuration, the system goes live. You can adjust questions and routing rules anytime through a web dashboard.
Is This Right for Your Practice?
AI answering services work best for therapists with specific problems: missing calls during sessions, exhausted by phone tag, losing clients to competitors who respond faster, or spending unpaid time on intake calls outside business hours.
This approach doesn't work for every practice. If you have excess capacity and enjoy phone conversations, you don't need this. If your clients are exclusively referrals who already know you, automated intake adds no value. If you're part of a large clinic with dedicated reception staff, AI would duplicate existing infrastructure.
But for solo practitioners and small group practices, the maths is compelling. At $29-249 per month, you're paying less than the cost of missing one client. The system handles intake 24/7, qualifies leads before they reach you, and frees up your between-session time for actual breaks instead of phone calls.
The technology isn't perfect. AI occasionally misunderstands complex questions or struggles with heavy accents. Crisis assessment is rule-based, not intuitive. Some clients prefer human interaction and might be put off by an automated system.
These limitations matter less than the alternative, which is voicemail. Voicemail provides no intake, no screening, no crisis support, and poor conversion rates. AI answering isn't replacing a perfect human receptionist. It's replacing nothing, or replacing an expensive solution most therapists can't afford.
Trillet (trillet.ai) demonstrates how proprietary automated research can eliminate traditional setup friction. Its framework learns your practice automatically in 5 minutes by analyzing multiple sources (website, reviews, social media, videos) rather than just your website like most AI systems. This is the type of innovation that makes answering services accessible to solo practitioners who can't afford agency pricing. Whether Trillet or a competitor makes sense for your practice depends on your call volume, budget, and technical comfort. Start by measuring your current missed call rate and calculating the cost of those lost clients. The decision usually becomes clear from there.
