Both audiences show up, and you should design for both. Many of these visits are the adult child comparing communities on a phone at eleven at night, and some are the older adult reading the site themselves. Neither of them goes hunting for a small bubble in the corner, which is the real objection. A Tars agent can run as the primary experience on the page instead, taking the place of the form, opening with a plain prompt like "tell me who you are looking for and I will find the right community," and it can be prompted differently on a pricing page than on a community page. We report engagement rate per page, so you can watch that number rather than hope.

No, and the reason is architectural rather than a promise. Pricing disclosure and care-level language do not run on model judgment. You define, once, what may be said about rates and availability at each community and at what point in the conversation, and the agent runs those steps as deterministic flows, exactly as written, on every channel. The agent answers from your approved content and nothing else, and anything outside that definition is handed to a person instead of improvised. Before launch you test it against your own worst-case questions and turn those into standing evaluators, so you see how it behaves before a family does.

It goes to a person immediately. The agent is operational: it answers questions about your communities, qualifies, and books tours. It does not give medical advice, does not assess whether someone needs a particular level of care, and does not promise that a community can meet a specific need. You define the triggers, for example a fall, a hospital discharge, or a caregiver in crisis, along with the exact wording you want used, and the agent stops the qualification questions and escalates with the full conversation attached.

One analytics home for the portfolio, and every number in it is clickable down to the conversation behind it. You get the funnel from visitors to conversations to booked tours, drop-off at each step, and conversions broken out by UTM source, campaign, and landing page with a conversion rate on each. You can filter to a single community or roll the portfolio up, compare any trend to the previous period, see which qualification answers families give and which they skip, and export any view. Booked tours are counted only when the booking actually went through, so the number in the dashboard is the number on the calendar.

This is the question that has killed more of these projects than any other, so we handle it before you commit rather than after. We run a pilot integration test against your actual CRM as part of evaluation, and you see the records land before there is a contract to sign. Tars connects to hundreds of tools plus APIs and webhooks for anything custom, and the record we write is structured, so lead source and qualification answers land in real fields your team can report on rather than buried in a notes box. In the worst case, where a system will not give us what we need, you still get the complete conversation record and the qualified inquiry delivered where your team can work it.

Not yet, and we are not going to invent one. Tars has 8+ years in conversational AI and runs this exact motion in adjacent industries: inquiry capture and appointment booking for a residential real estate brokerage that generated thousands of leads at a 62% response rate, and patient-facing agents in healthcare including International Medical Center, a large hospital group in Saudi Arabia, which automated more than 1,000,000 conversations on WhatsApp. What we will do is put live agents in front of you to use yourself, and run a scoped pilot on one region of your portfolio where the only thing that matters is whether tours went up. If you want to be the case study, that is a conversation worth having directly.

Tars is priced as a platform against the conversations it handles, so the arithmetic does not punish you for the size of your portfolio. Judge it against move-ins, not per community. Run your own rate and average length of stay against a handful of incremental move-ins across the whole portfolio in a year, and the platform has answered for itself. On building it, a chat box on a model is the easy part. The hard part is deterministic pricing and escalation flows your legal team will sign, evaluation that proves behavior on the worst inputs before a family sees it, one conversation that follows a family across web, SMS, WhatsApp, and email for months, tour booking on your counselors' real calendars, audit trails and compliance certifications, source-to-tour attribution, and your counselors taking over inside the same thread. That is what Tars has spent 8+ years on, and you deploy in weeks. Phone coverage is on our roadmap rather than shipped, and we would rather tell you that now than discover it in month three.