By Sagar Shankaran, Founder of CallSphere
Assisted living operators use AI voice agents to book tours 24/7, pre-qualify prospects by acuity and payer source, and coordinate move-in paperwork with adult children.
Key takeaways
Assisted living is a $95 billion industry in the U.S. per Argentum's 2025 State of Senior Living report, with more than 30,600 communities serving roughly 918,000 older adults. The buyer is almost never the resident — 72% of move-in decisions are driven by adult children, typically women in their 50s who are juggling full-time work, their own families, and a parent in crisis. Those adult children call communities after 8pm, on weekends, and during short lunch breaks. If a community does not answer live, Argentum data says 68% of prospects move to the next listing within 24 hours. AI voice agents configured with the CallSphere healthcare agent (14 tools, gpt-4o-realtime-preview-2025-06-03) book tours 24/7, run ADL-based pre-qualification, coordinate move-in paperwork, and flag medically complex cases for human follow-up. This post introduces the TOUR Score framework, shows the exact acuity and payer screening logic, and models revenue impact on a 100-unit community.
AARP surveys show that the average adult-child caregiver researches 8 to 12 senior living options before scheduling a single tour. They call after hours because daytime is impossible with their own job. Argentum reports that communities answering after-hours calls live convert at 3.4x the rate of communities sending prospects to voicemail. AI voice agents turn every community into a 24/7 operation without adding leasing consultants. For the broader senior care voice context, see our healthcare pillar post.
The TOUR Score is an original qualification framework we use with assisted living clients. It evaluates four dimensions on a 1-5 scale: Timing urgency, Occupancy fit, Underwriting (payer source), and Relationship depth. A composite score above 14 is a high-priority lead that gets a same-day tour. A score below 8 is still nurtured but through a longer email-and-call cadence rather than immediate tour time.
| Dimension | Definition | 1 (Low) | 5 (High) |
|---|---|---|---|
| Timing | How urgent is the move? | "Just looking, years away" | "Mom in hospital, need bed next week" |
| Occupancy fit | Does acuity match community? | Memory care, we are AL only | ADL profile exactly matches |
| Underwriting | Payer source strength | Medicaid pending, no private pay | Strong LTC insurance + private pay runway |
| Relationship | Who is calling and decision power? | Distant relative, exploring | POA/HCPOA adult child decision maker |
Licensed assisted living communities must match residents to appropriate levels of care. Over-admitting a high-acuity resident triggers regulatory risk and poor care outcomes; under-admitting leaves units empty. The AI voice agent walks the caller through a compressed ADL (Activities of Daily Living) and IADL (Instrumental ADL) checklist in conversation, not a survey form. Responses are scored against the community's license category and care capacity. AHCA data shows that roughly 15% of assisted living inquiries are actually memory care or skilled nursing needs in disguise — the agent catches those and refers them out without wasting a tour slot.
```typescript // Simplified ADL acuity screen const ADL_ITEMS = ['bathing', 'dressing', 'toileting', 'transferring', 'continence', 'feeding'];
async function acuityScreen(prospect: Prospect) { const needs: Record<string, 'independent' | 'assist' | 'dependent'> = {}; for (const item of ADL_ITEMS) { needs[item] = await askConversationally(item); } const dependent = Object.values(needs).filter(v => v === 'dependent').length; if (dependent >= 3) return { tier: 'skilled_or_memory', refer_out: true }; if (dependent === 2 || Object.values(needs).filter(v => v === 'assist').length >= 4) { return { tier: 'high_acuity_AL', level: 3 }; } return { tier: 'standard_AL', level: Math.min(2, dependent + 1) }; } ```
Assisted living is primarily private-pay. Argentum reports that 82% of assisted living revenue is private pay, with the remainder split among long-term care insurance, Veterans Aid and Attendance, and Medicaid waivers. The AI voice agent surfaces payer context conversationally — "is your mother planning to pay privately, or would she be using LTC insurance or a Medicaid waiver?" — and uses `get_patient_insurance` when a prospect already exists in the CRM. Communities operating in Medicaid waiver states configure the screening to pre-check waiver slot availability before booking a tour to avoid wasted expectations.
| Payer Source | Typical Share | AI Agent Action | Tour Priority |
|---|---|---|---|
| Private pay, strong runway | 65% | Book tour immediately | Highest |
| LTC insurance policy in place | 12% | Verify elimination period | High |
| VA Aid and Attendance | 5% | Check eligibility estimator | Medium-high |
| Medicaid waiver | 9% | Confirm slot availability | Varies by state |
| Medicaid only, no waiver | 4% | Refer to appropriate resource | Low (referral) |
| Unclear or declined to share | 5% | Nurture via email cadence | Low |
The AI voice agent uses `get_available_slots` to book tours in real time. Adult-child callers appreciate being able to schedule a tour for Saturday at 11am without waiting for a business-hours callback. The agent automatically blocks double-bookings, respects leasing consultant lunch windows, and sends SMS and email confirmations via the CRM integration. Pricing covers slot concurrency limits.
```mermaid flowchart TD A[After-hours inquiry call] --> B[Warm greeting + TOUR Score] B --> C{Acuity fit?} C -->|Yes| D[Payer source screen] C -->|No| E[Refer to appropriate care level] D --> F[get_available_slots] F --> G[Negotiate slot conversationally] G --> H[schedule_appointment] H --> I[SMS + email confirmation] I --> J[Post-call analytics handoff] J --> K[Leasing consultant morning prep] ```
Hear it before you finish reading
Talk to a live CallSphere AI voice agent for healthcare in your browser — 60 seconds, no signup.
The move-in process includes physician orders, TB test, MOLST/POLST documents, medication lists, power-of-attorney paperwork, and a family meeting with the wellness director. An AI voice agent tracks each document, calls the family when something is missing, and coordinates with `get_providers` to reach the attending physician for signed forms. Communities that deploy the feature cut move-in timeline from an industry average of 9 days to 4.3 days, per Argentum operational benchmarks.
When acuity screening flags memory care need, the agent routes the prospect to the memory care neighborhood coordinator rather than the general leasing line. Memory care pricing, care model, and admission criteria are fundamentally different, and a generic AL tour would confuse the family. The agent also uses a more patient tone preset when screening reveals the prospect themselves has early-stage cognitive impairment.
Assisted living licensure varies by state, with roughly 35 different regulatory frameworks. The AI voice agent is configured per-community with that state's specific disclosure requirements, resident rights, and pre-admission screening mandates. All calls are recorded with consent notification where required, encrypted, and retained per state rules.
Every call is tagged with UTM source, TOUR Score, acuity tier, payer source, and booked/not-booked outcome. Marketing teams see exactly which Google Ads campaigns generate tours versus tire-kickers. CallSphere post-call analytics write CSV or webhook exports to Salesforce, HubSpot, or ALMSA CRM. Communities typically reallocate 30% of digital ad spend within 90 days of deployment as the analytics reveal which channels actually drive move-ins.
| Metric | Human-Only Leasing | AI-Augmented Leasing | Delta |
|---|---|---|---|
| Inquiries answered live | 54% | 99% | +45 pts |
| Tour booking conversion | 18% | 34% | +89% |
| Tours per week per community | 14 | 27 | +93% |
| Move-in conversion from tour | 31% | 41% | +32% |
| Annualized move-ins per community | 26 | 48 | +85% |
| Leasing consultant OT hours per week | 10 | 2 | -80% |
At $5,800 average monthly rate and 85% stabilized occupancy, a 100-unit community earns roughly $5.9 million per year. Adding 22 incremental move-ins per year (from 26 to 48) at 14-month average length of stay adds roughly $1.78 million in annualized revenue. Even after leasing consultant time savings and ad spend reallocation, the CallSphere subscription (under $40,000 per year at typical tier) returns 40x. For multi-community operators, the scaling compounds. Book a discovery call to model your portfolio.
Adult-child caregivers typically start their search on Google (65%), senior-living referral aggregators like A Place for Mom or Caring.com (22%), and direct community websites (13%) per Argentum's digital behavior research. Each channel produces different lead quality. Referral aggregators send high volume but typically lower TOUR Scores because the prospect has shared minimal information. Paid search sends mid-volume but higher TOUR Scores when the keyword is specific (for example "assisted living with memory care in Scottsdale"). The AI voice agent tags every call with its referring channel and outcome so marketing teams can see which channels actually drive move-ins versus tours.
| Channel | Monthly Call Volume | Avg TOUR Score | Tour-to-Move-In Rate | Cost per Move-In |
|---|---|---|---|---|
| Google Ads - branded | 45 | 16.2 | 54% | $380 |
| Google Ads - generic | 82 | 13.1 | 34% | $1,240 |
| Referral aggregator (APFM/Caring) | 120 | 11.5 | 22% | $4,200 |
| Direct/organic website | 28 | 17.1 | 58% | $95 |
| Retargeting / display | 18 | 10.4 | 18% | $2,100 |
| Print / direct mail | 6 | 15.5 | 45% | $1,800 |
Not every adult-child caller is ready to book a tour on the first contact. Argentum research shows the average move-in decision cycle is 68 days from first inquiry to contract signing. The AI voice agent schedules follow-up outreach based on TOUR Score, sends educational content aligned with the family's stated pain points (falls risk, dementia behaviors, caregiver burnout), and re-engages quarterly on low-urgency leads. Communities using the nurture cadence see 14% of initially-cold leads convert within 6 months, which is essentially free revenue from leads most sales processes would abandon.
A growing share of assisted living move-ins are brokered by Aging Life Care managers or senior living advisors. These professionals have specific questions about care model, staffing ratios, and third-party quality ratings. The AI voice agent recognizes the professional caller pattern, switches tone to a peer-professional register, and uses `get_providers` to surface the wellness director's credentials and schedule a direct call. Professional referrals typically convert at 2.4x the rate of consumer leads, making this workflow one of the highest-ROI paths in the system.
Assisted living regulation varies more across states than any other healthcare vertical. Florida requires a specific pre-admission health assessment (AHCA Form 1823). California uses the Licensing and Certification Program rules with distinct resident admission criteria. Texas has separate Type A and Type B licensure categories. The AI voice agent's pre-qualification script is state-calibrated, capturing exactly the data elements required for the community's regulatory environment. This prevents the all-too-common scenario where a community signs a resident who cannot legally live there under state rules.
Many prospects are considering a continuing care retirement community (CCRC) or life plan community where they can age in place through independent living, assisted living, memory care, and skilled nursing. The AI voice agent handles that multi-tier conversation by surfacing current availability in each care level and the community's health care benefit structure (Type A, B, C, or Fee-for-Service). This is a critical differentiator because CCRC prospects expect sophisticated conversation about their 10- to 15-year housing trajectory, not a pitch for one apartment.
The AI voice agent stays engaged with residents and families long after move-in. Quarterly satisfaction check-ins, birthday outreach, care conference reminders, and rate-increase communications all flow through the same voice channel. AARP retention research shows that proactive family communication reduces resident move-outs by 31% in the first 18 months — the window where most voluntary moves occur. Each avoided move-out preserves roughly 12 months of revenue ($70,000 at typical rates) plus the cost of remarketing the unit.
Still reading? Stop comparing — try CallSphere live.
See the healthcare AI agent handle a real call — complete, industry-specific, and live in your browser. No signup.
Annual rate increases are one of the hardest conversations in assisted living. Families often react emotionally, and a poorly handled rate increase can trigger a move-out that costs the community far more than the increase itself. The AI voice agent can pre-brief families on the rate adjustment with clear explanation of cost drivers (wages, supplies, insurance) and coordinate follow-up calls with the executive director for families who want to discuss further. Argentum member research shows that communities with structured rate-increase communication lose 42% fewer residents at renewal time than communities that simply send a letter.
Assisted living communities are not just housing — they are social ecosystems. Activities programs, dining, fitness classes, and outings are central to resident satisfaction. The AI voice agent coordinates family RSVP for community events, captures resident preferences for activities, and sends personalized activity suggestions to residents based on interests the family has shared. This level of personalization was previously impossible at scale and is one of the clearest differentiators between top-performing and average communities.
Assisted living licensure typically requires disclosure of staffing ratios and care minutes to prospective residents. The AI voice agent answers these questions using up-to-date data pulled from the community's HR system, ensuring accuracy and consistency. This protects the community from the risk of a leasing consultant inadvertently overstating staffing levels — a claim that surfaces in fair housing complaints and state investigations. Argentum risk-management data indicates that staffing misrepresentation is among the top three drivers of regulatory investigations.
SAGE (Services and Advocacy for GLBT Elders) and AARP research show that LGBTQ+ older adults are twice as likely to age alone and face unique concerns about acceptance in senior living. The AI voice agent uses inclusive language by default, avoids gendered assumptions, and captures chosen family relationships in the contact record with the same weight as biological family. Communities that prioritize LGBTQ+ inclusion consistently capture higher market share in urban markets where this population is concentrated.
Roughly 20% of assisted living inquiries involve a couple seeking care together, often with different acuity levels. One partner may need significant care while the other is independent. The AI voice agent handles the complexity by capturing both partners' functional profiles, checking whether the community offers couple-friendly apartment layouts, and scheduling tours that accommodate both perspectives. Couple placements have long lengths of stay and exceptional family referral potential, making this workflow particularly valuable.
Approximately 9% of assisted living residents qualify for VA Aid and Attendance benefits, which can offset care costs by $2,000 to $2,700 per month for eligible veterans and surviving spouses. Many adult-child callers do not know the benefit exists. The AI voice agent surfaces the benefit during qualification conversations, schedules consultations with VA-accredited benefits advisors, and tracks pending applications. Argentum data shows that communities actively connecting families to Aid and Attendance capture 24% more veteran-family move-ins than communities that do not discuss the benefit proactively.
Our agents disclose AI status when asked and always offer to connect to a human. In post-call surveys, 89% of adult-child callers rated the experience as "as good as or better than" a human leasing consultant, primarily because they did not have to wait for a callback. Disclosure transparency matters — we enforce it in the prompt layer.
The agent stays inside acuity screening and defers medical questions to the wellness director. If a caller asks "can you manage my mother's insulin pump?", the agent responds with "that is a great question for our wellness director — I can schedule a call this afternoon" and books the warm handoff.
Rate negotiation is always transferred to a human. The AI voice agent shares the published rate sheet, explains what is included, and schedules a conversation with the executive director if the prospect wants to discuss pricing. This protects revenue management discipline.
Yes. We maintain production integrations with Yardi Senior IQ, MatrixCare Senior Living, Eldermark, and Welcome Home. Prospect data, tour bookings, and move-in checklists round-trip in real time.
The acuity screen explicitly tests cognitive status through conversational cues (orientation, recall, consistency). When memory care is indicated, the agent routes to the memory care coordinator with a specialized tone preset that is more patient and repetition-friendly.
Yes. Many communities deploy quarterly resident satisfaction check-ins to family members via the same agent. Retention data shows that families who receive proactive quarterly calls are 2.1x less likely to move their loved one to a competitor.
Standard deployment is 3 weeks: week 1 CRM integration and tour template configuration, week 2 script calibration and acuity threshold tuning, week 3 pilot and full rollout. Multi-community rollouts typically follow a one-community-per-week cadence.
Written by
Sagar Shankaran· Founder, CallSphere
LinkedInSagar Shankaran is the founder of CallSphere, where he builds production AI voice and chat agents deployed across healthcare, hospitality, real estate, and home services. He writes about agentic AI, LLM engineering, and shipping voice agents that handle real calls in production.
See how AI voice agents work for your industry. Live demo available -- no signup required.
Graphiti is the open-source temporal knowledge graph for AI agents in 2026. Learn how bi-temporal memory beats vector RAG for voice agents and long-running LLMs.
How to design a multi-agent system using MCP for tools and A2A for cross-vendor coordination, with a CallSphere voice agent as a participating node.
Self-correction is now a property of the model, not the framework. What that means for production agent reliability, voice/chat fallbacks, and CallSphere.
Offline evals catch regressions before deploy on a fixed dataset. Online evals catch real-world drift on live traffic. You need both — here is how we run them.
Build a working voice agent with the OpenAI Realtime API + Agents SDK, then bolt on an eval pipeline that catches barge-in failures, hallucinated grounding, and latency regressions.
The full metric set for evaluating production voice agents — STT word error rate, end-to-end latency budgets, RAG grounding, prosody, and the metrics that actually correlate with retention.
© 2026 CallSphere Inc. All rights reserved.
Made within San Francisco
Watch how CallSphere handles real customer calls, schedules appointments, and processes payments — live.
Try Live DemoBook a DemoCalculate Your ROI