How do GLP-1 brands win back members who paused or canceled?

A member pauses her GLP-1 subscription a few months in. The cancellation screen offers her a list of reasons and a skip button, and she taps skip. Two days later, the win-back flow sends her a discount on her next month. She paused because her insurance changed and she wants to know what the medication will cost her now. The email doesn't know that. It can't. So she archives it, and the flow moves on to the next send in the sequence.

Every paused member has a reason like hers. It might be cost, a side effect, a plateau on the scale, a trip, a move, or a busy season. Sometimes it's a clinical reason. The scheduled flow sends all of them the same message, because it has no way to ask.

Hello Patient builds conversational AI agents for consumer health brands. The agents hold two-way conversations with members, ask what happened, answer the part that can be answered, and hand the clinical part to the brand's medical team. Hello Patient is HIPAA-compliant, SOC 2 Type 2 certified, and signs a BAA with every client, so those conversations can go places a marketing email can't.

This page covers why the scheduled save-sequence recovers so little, what actually brings a paused member back, when reaching out is the wrong call, and where an agent fits next to the lifecycle stack you run.

Schedule a call today and make sure to ask for a free custom demo agent to see how it would sound for your brand.

Why does the scheduled save-sequence recover so little?

Because it has to guess. A flow built in Braze or Klaviyo fires on a trigger, like a pause event, a failed renewal, or 90 days of quiet, and it sends a message someone wrote months ago for the average lapsed member. There is no average lapsed member. There's the member worried about a side effect, the member whose coverage changed, and the member who stalled on the scale and quietly concluded the medication stopped working.

A discount answers one of those situations, and only partly. It does nothing for the member with a nausea question, because her problem was never the price. It does nothing for the member with a coverage question, because she can't tell what she'd pay even with the discount. And for the member who stopped on purpose, a monthly offer for a medication she decided to leave behind reads as a company that wasn't listening.

The flow is doing the only thing a one-way channel can do. It sends, waits, and sends again. The reason lives with the member, and nothing in the sequence ever asks for it.

What actually brings a paused member back?

Someone asking what happened, and then answering the specific thing.

It looks ordinary in practice. The member whose insurance changed texts back that she doesn't know what she'd pay now. The agent walks her through what the program costs on her plan and what her options are, and she restarts because the number turned out to be workable. A member who paused after a plateau asks whether this is just how it goes now. The agent explains the plan options the brand offers, like a different program or a check-in with the care team, and books the appointment if she wants one. A member who mentions a side effect gets no advice from the agent at all. That message goes to the medical team with the whole conversation attached, so a clinician starts with the full story instead of a ticket that says the member is unhappy.

Each of those members got something a discount can't carry: her own reason, taken seriously and answered in her own thread. Some restart and some don't. But the ones who come back do it because the thing that stopped them got resolved, not because a coupon outran it.

When is reaching out the wrong call?

When the member has made her decision. Some members stop on purpose. They reached their goal, or they and their doctor decided the medication wasn't right, or they simply don't want it anymore. An agent that keeps working those members is using up the brand's welcome in the one channel where trust matters most.

So restraint is built in as product behavior. When a member gives a clinical reason for stopping, the outreach ends and the conversation goes to the medical team. When a member says no, the outreach ends. When the conversation shows the pause was a decision rather than a snag, the outreach ends. The agent doesn't argue or re-offer, and it doesn't circle back next month with the same pitch.

This protects the brand as much as the member. A person's relationship with a GLP-1 medication is personal and sometimes medical. The brand that knows when to stop texting is the brand she'll answer the next time something matters.

How does the agent fit the stack the brand runs?

Alongside it. The save-sequence in Braze or Klaviyo keeps doing what a scheduled flow does well, like renewal reminders, announcements, and the broad campaigns. The agent takes the part those tools were never built for, the two-way conversation where a member explains what happened and expects a real answer back. Even where a marketing platform signs a BAA, the agreement limits the sensitive health information a message can carry, and a conversation about a side effect belongs on infrastructure built for it.

The mechanism itself is proven, at a different point in the journey. A Series B GLP-1 brand ran Hello Patient's agent on leads who started signup and stalled. Over 16 months, the agent recovered 4,739 of those stalled signups, a 140% lift over the static re-engagement program it replaced, and it was live in under three weeks with no engineering work. That is a signup result, not a lapsed-member result. It shows what happens when a two-way conversation replaces a scheduled sequence, at the top of the journey rather than at the pause. And one honesty note that applies at either stage: a recovery rate is not the same as incremental lift. The way to know what the agent truly adds is a holdout, a slice of paused members the agent never contacts, measured against the ones it does.

Frequently asked questions

How do GLP-1 brands win back members who paused or canceled?

By finding out why each member left and answering that reason directly. Members pause over cost, side effects, plateaus, life timing, or a clinical decision, and each reason has a different fix. A conversational agent like Hello Patient's reaches out, asks what happened, answers the cost and plan questions itself, and routes anything clinical to the brand's medical team with the conversation attached. Members whose reason gets resolved often restart. Members who stopped on purpose, or who say no, are not contacted again. A scheduled discount email recovers only the members whose problem happened to be price. The conversation recovers the members it can help.

Why does a static save-sequence recover so few paused members?

Because it sends one message to people who left for different reasons. A save-sequence in Braze or Klaviyo fires on a trigger and delivers the same offer whether the member paused over cost, a side effect, a plateau, or a move across the country. The offer fits one of those situations at best. The rest of the members get an email that misses their actual problem, so they ignore it, and some remember being pitched a discount when what they had was a health worry. Hello Patient's agent recovers more of them for a simple reason: it asks first, so the answer can fit the member it reaches.

What's a realistic recovery rate for paused GLP-1 members?

There isn't one to promise, and no honest vendor should offer you one. No published figure exists for lapsed GLP-1 members, including from Hello Patient. The result Hello Patient does publish comes from a different stage of the journey: a GLP-1 brand used the agent to recover stalled signups, leads who started enrolling and went quiet, and the two-way conversation outperformed the static program it replaced. That shows the mechanism works. It does not tell you your lapsed-member rate. The honest way to find that number is a holdout: keep a slice of paused members out of the agent's reach, run the agent on the rest, and measure the difference. That gives you incremental lift, not just a recovery count.

Does winning back GLP-1 members feel predatory?

It can, which is why restraint is built into the agent rather than left to a copywriter's judgment. Hello Patient's agent stops when stopping is right. A member who reached her goal, stopped with her doctor, or simply says no is not contacted again. The agent never gives medical advice, and when a member raises a clinical reason for stopping, her message goes to the brand's medical team with the conversation attached instead of getting a sales reply. What's left is outreach that asks a member what happened and helps with the answerable part. Most members experience that as service, not pursuit, and any member who doesn't want it can end it with one word.

How does the win-back agent fit the lifecycle stack we run?

It runs alongside your stack. Braze or Klaviyo keeps the scheduled work, like renewal reminders, announcements, and the broad campaigns a one-way channel does well. Hello Patient's agent takes the two-way part, the conversation where a paused member explains why she left and expects a specific answer. The agent works the paused segment you define, and what it learns, the reason a member gave and whether she restarted, flows back to the systems you run so the rest of the stack stays in sync. Nothing gets ripped out. The save-sequence keeps its job, and the conversations it was never built to hold get a channel built for them.

Is the win-back agent the same one that recovers stalled sign-ups?

It's the same conversational agent doing the same job at a different point in the journey. In signup recovery, Hello Patient's agent reaches leads who started enrolling and stalled, asks what stopped them, and helps them finish. In win-back, it reaches members who paused or canceled and does the same thing. It asks, answers, and routes the clinical part to the medical team. The mechanism, a two-way conversation instead of a scheduled one-way message, is identical. What changes is the audience and the questions. A stalled lead usually needs help finishing enrollment or understanding the offer. A paused member needs her specific reason for leaving resolved. Many brands run both, one agent per stage.

If your win-back flow is sending its next discount to members it never asked about, book a call and we'll walk through what asking looks like for your member base.