All use cases

Convert

Warm referral from free user to paid clinical consult

A free user asks the weight-medication chat a question only a clinician should really answer — the agent helps with what it can, then opens the door to the paid consult instead of leaving the question to evaporate.

Get aheadNext best thingBehavioral triggerChatWhatsAppIn-appPushGLP-1 & PeptidesHealth & Telehealth

What it does

GLP-1 programs run free chat surfaces that absorb a steady stream of medication questions, and a large share of that support traffic is exactly the kind of clinical question with no paid route attached today. The user gets a general answer and leaves; the paid clinical consult that would actually serve them sits one tap away, never offered, and the free surface stays a pure cost. The agent recognizes when a free or unpaid user's medication question carries genuine program-fit intent, answers what is safe to answer in the free scope, and warmly routes them into the paid consult — request, intake, secure payment, then connection to a clinician — turning unrouted medical-question volume into started clinical care.

How it works

  1. 1
    Trigger. a free or unpaid user's message in the chat, WhatsApp, in-app thread, or a push-prompted return shows a medication question that warrants clinical input — a side-effect worry, a "should I be on this" question, a dose or eligibility question beyond the free scope.
  2. 2
    Decision. clinical red-flag detection runs first, so anything urgent escalates to a clinician rather than being routed to a paid funnel. Then the agent reads memory — what this user has asked before, whether they have already declined the offer — and a quality judge gates the referral for tone: warm and optional, never pressuring, and only where the consult is genuinely the better answer.
  3. 3
    Action. the agent answers within the free educational scope, then offers the paid clinical consult as the honest next step in the same thread — you request it, fill a short intake, complete payment on the program's secure step, and a clinician picks it up.
  4. 4
    Follow-through. a decline writes to memory and sets a cool-down so the next free question is not met with another pitch; the payment completes on the program's own hosted step, never in the conversation; conversion from referral is measured against a holdout so the lift is real, not assumed.
Example
CustomerThe nausea's been rough since I started the medication and I'm not sure if I should keep going or change something. Any advice?
AgentEarly nausea is common and there are gentle things that help — smaller meals, hydration, eating slowly. Whether to adjust anything is a clinician's call, not something I can advise. The quickest way to get a real answer is a consult: you send the question, add a couple of details, and a clinician reviews and replies with a plan. Want me to set that up?
CustomerYeah, how do I start it?
AgentYou answer a few questions right here, complete a quick secure checkout, and a clinician takes it from there. I'll hand you to the secure step now and bring you back to this thread once you're connected.

Configuration

How the agent is wired for this use case.

Triggera free-tier user's message showing a clinical-input-worthy medication question with program-fit intent — detected in the chat, WhatsApp, in-app thread, or on a push-prompted return.
Tools & actions
  • Messaging channel · read the inbound message, answer within the free educational scope, and offer the consult referral in the same thread.
  • App backend · read the user's tier and entitlement state to confirm the user is unpaid and the paid consult applies.
  • Knowledge base · source the general, non-clinical guidance the agent shares before referring.
  • Scheduling · open the consult request and the short intake when the user accepts.
  • Hosted payment step (program-owned) · receive the hand-off to take the consult payment; the agent never takes payment itself.
  • CRM · record the referral, the outcome, and any decline cool-down for measurement against a holdout.
Autonomyanswering and referral composition run unattended, judge-gated for warm, optional tone on every referral, but only after clinical red-flag detection clears. Clinical = detect and escalate, never diagnose or advise a dose. A care-first policy means the free answer is never withheld to force the upgrade, and the consult payment always routes to the program's hosted step.
Channelschat · whatsapp · in-app · push
Escalationanything clinically urgent or a red-flag symptom escalates to a clinician instead of being referred to the paid funnel; the consult payment hands off to the program's hosted step; a decline sets a cool-down before the next referral.

What you need

The inputs this use case runs on. Your channels stay yours; the agent supplies the judgment.

Signals

chat, WhatsApp, in-app, and push messages from free-tier users, plus intent classification that separates a clinical-input-worthy question from one the free scope can fully answer.

Data

user tier and entitlement state, consent state, memory of prior questions and declined referrals, the paid-consult catalogue and its intake fields.

Guardrails

clinical red-flag detection runs before any referral; judge gating for tone on every referral; decline cool-downs and frequency caps; care-first policy so the free answer is never withheld to force the upgrade; the agent gives no medical or dosage advice and routes clinical decisions to a clinician.

Metrics it moves

  • conversion-rateup on free-to-paid, because the consult is offered exactly when the user's medication question already justifies clinical input.
  • revenue-per-conversationup, as a large share of free medical-question traffic that had no paid route starts producing consults.
  • csatheld or up, because the referral is gated to moments where a clinician is genuinely the better answer, not pushed on every question.

Related use cases

Upsell & Cross-sell

Free-to-paid nudge in the care chat

the cross-vertical parent this specializes for the GLP-1 paid consult

Prevent / Care

Clinical red-flag detection and safe escalation

the safety gate that runs before any referral

Prevent / Care

Doctor copilot — pre-drafted clinical reply with citations

what the clinician picks up once the consult is paid

See it on your own customer journey

Bring one drop-off, one churn cliff, or one silent segment. We will show you what a proactive agent with memory and judgment does with it.

Book a demo