Guides
Build a Thoughtly workflow that follows up with first-party senior living inquiries, books verified community tours, hands sensitive questions to people, and measures the path to move-in.
Last updated
Senior living inquiry follow-up should turn a first-party family or prospective-resident inquiry into a useful next step: a verified community tour, a scheduled conversation with a sales counselor, or a clean human handoffHuman handoffThe moment an AI agent transfers context, call details, and the next step to a human rep, licensed specialist, or support team.. The job is not to run a generic calling sequence. It is to preserve the context behind a high-consideration decision while keeping clinical judgment, care eligibilityEligibilityThe fit criteria that determine whether a prospect can move forward, such as service area, insurance coverage, loan type, location, age, or program requirements., pricing commitments, and resident-specific health questions with trained people.
This guide shows how to build that workflowWorkflowAn automated, multi-step process — usually triggered by an event (form fill, new lead) and orchestrating one or more voice / SMS / email actions. in Thoughtly. It covers the triggerTriggerThe event or condition that starts an automated workflow, such as a new lead, missed call, CRM status change, calendar booking, or completed call., contact model, conversation flowConversation flowThe designed path an AI agent follows through greeting, intent capture, qualification, routing, objections, fallback, and next-step confirmation., tour scheduling, multichannel follow-up, CRM write-backCRM write-backUpdating the CRM after an interaction with call outcomes, transcripts, qualification answers, notes, appointments, dispositions, and next-step fields., and measurement. For the earlier platform-selection decision, use our buyer guide to AI voiceAI voiceAn artificially generated, natural-sounding voice produced by a TTS model. Thoughtly supports a library of AI voices and brand-specific cloning. agents for senior living. A comparison page helps you choose a platform; this implementation guide defines what the platform should actually do after an inquiry arrives.
Thoughtly's point of view is simple: speed matters, but a fast call without the right community, prospect, family contact, tour inventory, and stop conditions just creates faster confusion. Build the record and operating policy first. Then let the agent cover every eligible inquiry consistently.
A senior living inquiry often involves more than one person. The person submitting the form may be the prospective resident, an adult child, another family member, or a professional referral source. The requested community may not be the community that ultimately fits. Tour availability can change. Questions can move quickly from amenities and location into medication, mobility, memory care, or other clinical territory.
That combination makes a generic lead form a weak source of truth. The workflow must answer three separate questions: who should be contacted, what sales context is safe and useful for the conversation, and which questions require a person. The strongest design does not ask the AI agent to imitate an admissions director. It gives the agent a narrow, repeatable job and makes the human handoff excellent.
Thoughtly Automations provide the deterministic layer for triggers, conditions, calls, messages, and system updates. The Agent Builder handles the live conversation with nodes, variables, outcomes, actions, and transfers. The senior living operator supplies the policy: eligible inquiry sources, approved talking points, tour inventory, routing rulesRouting rulesLogic that determines where a lead, call, appointment, or task should go based on criteria such as location, product, urgency, language, licensing, or availability., and the line between sales intake and clinical judgment.
| Stage | Thoughtly or system action | Required output | Human boundary |
|---|---|---|---|
| 1. Inquiry received | CRM, form, scheduler, or webhook sends a unique event | Contact, inquiry ID, source, permission, community interest | Reject duplicates, suppressed contacts, and unsupported sources |
| 2. Contact prepared | Create or update the Thoughtly Contact | Stable person record and limited persistent attributes | Do not merge different family members into one identity |
| 3. Eligibility checked | Conditions validate phone, permission, ownership, and timing | Eligible, delayed, or stopped disposition | A failed policy check never reaches the call step |
| 4. Conversation started | Call Contact with inquiry-specific metadata | Identity, relationship, community, timing, and next-step intent | No diagnosis, care-level determination, clinical advice, or guarantee |
| 5. Tour or handoff completed | Scheduling Action, warm transfer, or counselor task | Verified appointment or named owner with context | People own care, pricing exceptions, contracts, and sensitive questions |
| 6. Result written back | On Call Completed updates CRM or middleware | Controlled disposition, summary, fields, owner, next action | Failed writes create a visible exception queue |
| 7. Follow-up continued | Permitted SMS, email, or scheduled callback | Confirmation, reply, reschedule, or clean closure | Stop requests suppress future contact across channels |
The useful unit of work is not a dial. It is a state change that the sales team can trust. Each run should leave the inquiry closer to a tour, a counselor conversation, or a clear stop reason.
Start with a narrow event such as a new website inquiry for a named community, a callback request, a referral submitted through an approved channel, or an existing inquiry whose CRMCRMThe system of record for leads, contacts, deals, and activity. Thoughtly reads from and writes to your CRM continuously. stage makes a follow-up appropriate. Do not quietly expand the workflow to purchased lists or scraped family contacts. The consent, provenance, and expected seller matter more than the fact that a phone number exists.
Thoughtly supports native application triggers and incoming webhooks. Its trigger documentation recommends testing in Draft mode with real sample payloads, mapping only observed fields, and planning for idempotency because external systems may retry events. Those are not implementation niceties. They are how you prevent one family inquiry from creating multiple calls and conflicting owner records.
Require these fields before a call can be attempted:
The gate should stop a run when the event ID was already processed, the record is suppressed, required permission evidence is missing, the phone is invalid, or a human already owns an active conversation. Thoughtly's Automation Actions documentation supports Conditions, contact upserts, calls, messaging, scheduling, and external system steps that can implement this gate without asking the conversational agent to police the database.
The data model is the quiet difference between a senior living workflow and a generic follow-up bot. A family member can ask about two communities. Two siblings can ask about one prospective resident. One inquiry can change from independent living to a request for a trained care discussion. Flatten those facts into one contact record and the next call can easily use stale or mismatched context.
| Data layer | Examples | Where it belongs | Why |
|---|---|---|---|
| Persistent contact attributes | preferred_name, preferred_language, relationship_to_prospect, assigned_market | Thoughtly Contact or CRM person record | Reusable across future conversations |
| Inquiry metadata | inquiry_id, community_id, source, submitted_at, requested_callback_window | Metadata on this call or linked inquiry object | Belongs to one event, not permanently to the person |
| Conversation variables | identity_confirmed, tour_interest, move_timing, requested_handoff, selected_slot | Agent Builder variables | Captured or confirmed during the live conversation |
| Controlled outcomes | tour_booked, counselor_transfer, callback_scheduled, not_ready, wrong_person, stop_requested, needs_human_review | CRM disposition or Thoughtly outcome | Makes follow-up and reporting deterministic |
Thoughtly distinguishes persistent Attributes from per-call Metadata. Use Attributes for facts you expect to reuse, such as preferred language or assigned market. Use Metadata for the active inquiry, such as community ID, form source, and appointment campaign. Use Agent Builder variables for facts confirmed in the conversation. This keeps the agent from treating yesterday's inquiry as today's truth.
Do not use the prospective resident's name as the only matching key. Upsert the actual contact who supplied the phone number, retain the stated relationship as data, and link the inquiry to the prospect or household in the CRM. If identity or authority is uncertain, record the uncertainty and route it. A fluent conversation is not proof that the person can receive resident-specific information.
In Agent Builder, start with a short disclosure and identity check, then move through no more qualification than the sales team truly needs to route or schedule. The Agent Builder workflow processes variables before outcomes and actions, which is useful here: confirm who is speaking, extract the routing facts, choose the outcome, then run the scheduler or handoff action.
The agent should not interrogate the family. Every question must earn a routing decision, a booking decision, or better context for the person taking over. If a field does not change the next action, remove it from the call.
A senior living sales workflow should not determine a care level, interpret symptoms, advise on medication, promise clinical services, decide whether a resident is appropriate for a program, quote a binding price, or represent that space is guaranteed. Those questions should trigger a warm handoff, a scheduled counselor call, or a review task with the exact topic captured.
Not every senior living operator is a HIPAAHIPAAThe US health privacy law that governs protected health information. Healthcare voice and SMS workflows must handle PHI with appropriate safeguards. covered entity, and not every sales inquiry contains protected health information. Where a covered entity or business associate does handle PHIPHIProtected Health Information is individually identifiable health data regulated under HIPAA, including demographics, medical records, and payment information linked to a specific person., HHS explains that covered entities and business associates have specific Privacy and Security Rule obligations, and the minimum necessary standard generally requires limiting PHI to what is needed for the purpose. The practical design choice is to keep the automated sales workflow on non-clinical routing facts and send sensitive detail to an approved human process. Privacy and legal teams should approve the actual data boundary before launch.
Do not let the model invent a time or read from a pasted schedule. Use a live scheduling integration or an authoritative availability service. Thoughtly's Scheduling Guide describes a reliable pattern: fetch available times, capture the selected slot and timezone, run the booking Action, then branch on the booking result. Confirm only when the action returns a confirmed status; otherwise offer another time or route to a person.
A production booking should write at least these values:
Use a live transfer when the caller asks for a person, the question crosses the approved boundary, timing is urgent, or the inquiry is ready for a high-value counselor conversation. The transfer should carry identity, relationship, community interest, timing, questions raised, and what the agent already confirmed. Our guide to warm transfers with context covers the reusable mechanics; the senior living implementation adds the family-role, community, and clinical-stop context the counselor needs.
A transfer without context merely relocates the repetition. The handoff is successful when the human can begin with the caller's actual decision, not ask them to start over.
If the call does not connect, the next message should continue the same inquiry. A short SMS can identify the community, acknowledge the information request, and offer a reply path or approved booking link. Email can carry directions, a tour confirmation, or a recap. A scheduled callback should preserve the original inquiry metadata and the latest disposition.
Thoughtly supports SMS steps in Automations and per-contact history through Call Contact. Its Actions documentation also recommends using Contacts when history and persistent attributes matter. That is the right default for this workflow because families rarely resolve a senior living decision in one touch.
Permission and suppression rules belong before every channel action. The FCC's ruling on AI-generated voices states that AI-generated voices fall within TCPATCPAUS federal law governing telemarketing calls and SMS. Thoughtly enforces consent capture, time-of-day windows, and DNC scrubbing automatically. restrictions on artificial or prerecorded voice calls and discusses consent, identification, and opt-outOpt-outA recipient’s request to stop receiving calls or messages. Compliant systems must capture opt-outs and suppress future outreach where required. requirements. The FTC's Telemarketing Sales Rule guidance also describes entity-specific do-not-call duties and the limited inquiry-based established business relationship for live calls. Automated voice, text, state law, and the details of the inquiry can change the analysis, so counsel should approve the launch population, disclosures, hours, and suppression process.
The operating rule is stricter than 'keep trying.' Stop on an opt-out, wrong person, deceased prospect notice, sensitive complaint, legal concern, uncertain permission, or human ownership. A clean stop is a successful system behavior.
The transcriptTranscriptThe text record of a voice conversation, used for review, training, compliance audit, and search. is useful evidence, but it should not become the sales team's workflow. After the call, use a small disposition vocabulary, structured fields, a concise summary, and one explicit next action. Thoughtly recommends the On Call Completed trigger for post-call workflows. When a native integration is unavailable, webhooks can send call results to an external CRM or database.
For every disposition, write the inquiry ID, community, contact ID, event time, owner, last attempt, next action, and source. If a CRM update fails, route that failure to an exception queue instead of marking the workflow complete. Silent write failures are how teams end up calling families who already booked or opted out.
Test each node in text, then test the live voice experience with internal numbers. Thoughtly's Agent Builder documentation recommends validating logic with Test Agent before using Call Me for voice polish. Automation steps should remain in Draft until sample payloads, outputs, and external writes are verified.
A senior living pilot should include at least these edge cases:
Measure the funnel by source and community: eligible inquiries, contact rateContact rateThe percentage of inbound leads your team actually reaches by phone. Most B2C teams hover around 25%; Thoughtly typically delivers 90%+., meaningful conversations, tours booked, tours completed, counselor conversations, deposits, move-ins, and days to move-in. Track wrong-person events, duplicate suppression, sensitive-question handoffs, booking failures, and failed CRM writes as quality metrics. Use our conversion-lift measurement framework to compare the workflow with a valid human-only or prior-process baseline. Call volume is an input, not the business result.
Yes, when it reads from verified availability and the scheduling action confirms the booking. The agent should not invent times or promise availability from a static prompt. If the scheduler fails or the requested tour needs special handling, route the inquiry to a sales counselor.
Ask only what changes routing or the next step: who is speaking, relationship to the prospective resident, community interest, general timing, preferred contact method, and whether they want a tour or counselor conversation. Do not collect sensitive clinical detail merely because it might be useful later. Let trained people handle consequential care questions.
Create or match the actual person reached, store the stated relationship, and link the inquiry to the prospective resident or household in the CRM. Do not collapse two family members into one contact record. Share resident-specific information only under the operator's approved privacy and authorization rules.
Use the channel the person asked for when it is available and permitted. Voice is strong for a high-consideration conversation, while SMS and email are useful for confirmations, directions, booking links, and continuity after a missed call. The channels should share one inquiry state rather than launch unrelated sequences.
Transfer or schedule a human when the caller asks for one, raises a clinical or privacy-sensitive question, needs a pricing or contract exception, disputes prior information, has an urgent timeline, or is ready for a substantive counselor conversation. The handoff should include the confirmed context and the unresolved question.
Track completed tours, counselor conversations, deposits, move-ins, and time through the funnel by source and community. Also track safety and data quality: duplicate events stopped, opt-outs honored, sensitive questions routed, bookings confirmed, and CRM writes completed. A workflow that produces more calls but less trustworthy pipeline is not an improvement.
Build the smallest senior living inquiry follow-up workflow that can create a trusted tour, counselor conversation, or stop disposition. Keep the event contract strict, the data model explicit, the agent's questions narrow, the scheduler authoritative, and the human boundary obvious. Thoughtly is strongest here when it works as the conversion layer around the sales team: covering every eligible inquiry, carrying context across channels, and handing people conversations that are ready for judgment.
That is the difference between an automated calling demo and a production revenue workflow. The demo talks. The production system knows why it is calling, what it may decide, where it must stop, and which business outcome it is responsible for moving.
