How do I cut subscription churn at my supplement brand?

Why do supplement subscribers cancel after the first month?

The pattern sounds like this. The first order sells fine. The subscription take rate looks healthy. Then a large share of subscribers are gone before the second or third shipment. New subscribers keep arriving, the active count barely moves, and every month you pay Meta to replace the people you lost last month.

It is not just you. Recharge published order-sequence retention for roughly 1,800 supplement brands on its platform, following cohorts from July 2025 to June 2026. Only 86.6% of first-time supplement subscribers completed a first reorder, the lowest of the five consumable verticals Recharge compared, and 33.8% reached the third reorder (Recharge).

Two things stand out. Supplements lose more people before the second order than the four other categories Recharge measured, which kept 92.9% to 99.3% of first-time buyers to the first reorder. And supplement subscribers who get through the early renewals hold on about as well as any consumable category deeper in the lifecycle. (Recharge notes the 12-month figure is right-censored: recent subscribers have not had time to get there yet.)

So the work sits in the first renewals, starting with the second charge. That charge is the first time the customer pays without having just decided to buy. The first-order discount may be gone. The first bottle may not be finished. The customer may not remember choosing a subscription at all. Each of those is a different problem with a different fix, which is why "churn is high" is not a diagnosis. Left alone, early churn like this is one of the quiet causes of a growth plateau.

Is your churn one problem or four?

Before fixing anything, split the losses. A subscription can end in four ways, and each one shows up in a different report.

The third row is bigger than it looks. Recharge, for example, automatically cancels the subscriptions in its failed payment recovery strategy once every retry has failed, with the reason "Failed Payment flow max retries" (Recharge Help Center). Those customers did not decide to leave. Their card did. If your churn report lumps them in with people who chose to cancel, every fix gets aimed at the wrong channel.

Pull the last 90 days of cancellations with the reason and the channel or user that made the change. Sort them into the four buckets, then by subscriber age: first renewal, second, third, later. That one table tells you where the next month of work should go.

Who actually processes your cancellations?

This is the question most churn reviews skip. Brands build a careful cancel flow in their subscription app, watch the save rate, and assume it covers everyone. It only covers the people who cancel inside it.

Customers do not only cancel in the portal. They reply "please cancel" to a shipping email. They open a chat. They text. They call. They ask an AI support agent. In every one of those cases a person or a bot processes the cancel by hand, and the save offer you designed never appears.

Recharge's own documentation shows what that means on its platform:

• Cancellations made in the merchant portal, where support agents work, are not reflected in Cancellation Prevention analytics (Recharge Help Center).
• The cancel reasons an agent picks in the merchant portal default to the legacy portal list and do not automatically match the reasons in your Cancellation Prevention flow (same source).
• If you turn off portal cancellation or require a minimum number of charges, the portal shows your contact email instead of a cancel button, and customers must contact you to cancel (Recharge Help Center). Every one of those cancels then runs through support.
• One-click cancellation, when enabled, lets customers skip the cancellation survey (Recharge Help Center).

So a healthy save rate can sit right next to a falling subscriber count. The dashboard measures one door while customers leave through the others. Read the save rate itself with care too: Recharge counts a customer who leaves the exit survey without canceling as saved (Recharge Help Center).

In that example the cancel flow does its job and still only reaches 42% of the people leaving. Polishing the flow would barely move the total. Getting an offer in front of the other 58% would.

How do you put a save offer in every cancel channel?

Start with the menu. Recharge's Cancellation Prevention flows can offer a skip, a delay, a delay with a frequency change, a product swap, a one-time discount or a free gift, plus combinations such as delay and swap (Recharge Help Center). A pause step can also appear before the cancel survey (Recharge Help Center). Supplement subscribers already use skips: Recharge reports they skip 6.9% of renewal cycles, the highest skip rate of the verticals it compared (Recharge).

One quiet failure to check first: Recharge hides a cancel reason from customers when the action it offers (a skip, a swap, a discount code) is switched off in your customer portal settings (same Recharge source). A reason you built can be invisible.

Then put the same menu everywhere a cancel can happen:

• Support macros. One macro per top cancel reason, offering the matching option in a sentence or two, and confirming the cancel in the next reply if the customer repeats the request.
• Chat and SMS. The same options as buttons or short replies. Recharge's own AI SMS assistant, Concierge, treats a request to cancel an order as a charge delay once the customer confirms, and sends subscription cancellation requests to the customer portal (Recharge Help Center).
• AI support agents. If a helpdesk bot has permission to edit subscriptions, test what it does when someone types "cancel". Check whether it offers anything first, whether it logs a reason, and whether it cancels when asked twice.
• Phone. A short script with the same options and a clear path to cancel.

Why log a reason on every cancel?

Because the reason is the only part of a cancellation you can act on. "Too much product" is a cadence problem. "Too expensive" at the second charge is often a price step problem. "Not sure it's working" points back at the ad, the product page and the first weeks after purchase. Without the reason, all three blur into one churn number.

• Agents log a reason on every manual cancel. Make it a required field in the helpdesk or subscription admin, using the same list as the self-serve flow. On Recharge, that means updating the merchant portal reasons in your Cancellation Prevention settings so the two lists match (Recharge Help Center).
• Customers are asked, not forced. For subscriptions bought online, California requires online cancellation without further steps that obstruct or delay it (California BPC 17602). A short survey is one thing. A required questionnaire in front of the cancel button is another. Ask counsel where your flow sits.
• Keep "other" small. If "other" is your biggest reason, the list is wrong. Read the free-text answers and add the reasons that are missing.
• Read reasons by subscriber age and product. Recharge's reports break exit survey answers down by SKU (Recharge Help Center). A reason that spikes at the second charge on one product is a different problem from one spread evenly across a year.

Are failed payments quietly part of your churn?

Yes, and they are the cheapest churn to win back, because the customer never decided to leave. Recharge says that on average 7% of all recurring charges fail on the first attempt (Recharge's own figure) (Recharge).

The split that matters is soft versus hard declines. Stripe defines a hard decline as one where the issuing bank has rejected the transaction and it cannot be retried, such as a lost, stolen or pickup card, and only executes retries on those once a new payment method is added (Stripe Docs). Other declines are temporary: for an unreachable issuer, Stripe's guidance is to attempt the payment again (Stripe Docs), and Recharge retries insufficient funds declines (Recharge Help Center).

Here is the trap on Recharge. Charges that fail with a hard-decline code do not enter the Failed Payment Recovery strategy, are not retried automatically, and do not trigger the standard card declined email (Recharge Help Center). Recharge offers a separate Card Declined email for hard declines, and it is disabled by default (Recharge Help Center). Recharge also classifies some codes more strictly than you might expect: DO_NOT_HONOR is on its hard-decline list. If nobody switched that email on, a subscriber who froze or replaced a card may hear nothing at all.

More retries are not the answer either. Stripe notes that card networks limit how many times you can reattempt a single charge, recommends a maximum of eight retries for charges that permit them, and warns that issuers may see extra retries as potential fraud (Stripe Docs).

Card updaters fix part of the problem before it starts. Shopify Payments works with major card networks to update saved card details when customers get new cards (Shopify Help Center), and Stripe does the same, with wide support for US-issued cards (Stripe Docs). The gateway matters: Recharge does not automatically retry expired cards for stores using Stripe, Braintree or Authorize.net (Recharge Help Center).

Four failed-payment checks

• Is a card-update message on for hard declines, and does it link straight to a page where the card can be changed?
• What share of last quarter's cancels carry a max-retries reason instead of a customer reason?
• How long is your recovery window? Recharge's default automatic recovery period is 120 days after the original charge date (Recharge Help Center).
• Is a third-party dunning tool also running? Recharge says to disable third-party dunning before activating its own strategy (same source).

How many of your active subscriptions will never charge again?

This one hides in plain sight. A subscription can be marked active and have no next charge date. Recharge's help center says that when an active subscriber has no upcoming orders, there is likely an issue processing their order, and the error shows on the customer's profile and in the charge errors list (Recharge Help Center).

Some errors clear on their own. Others never will without someone touching them. Recharge notes that charge errors that must be retried by hand stay in an error state until you retry them, and are not automatically canceled (Recharge Help Center). A variant deleted from the catalog, a shipping or tax setup problem, an address the system rejects: each can leave a willing customer in limbo.

These subscribers never show up as churn, and they never pay. Because they still count as active, they can also flatter your churn rate. Recharge calculates churn as churned subscriptions divided by average daily active subscriptions (Recharge Help Center), so every stuck subscription sits in the denominator.

Does the renewal price surprise your subscribers?

Many supplement brands win subscribers with a first-order discount that is bigger than the recurring one. The ad says 40% off, checkout charges the discounted price, and a month later the card is charged a number the customer never really saw. The ones who leave over it leave at the second charge, because that is when they find out.

Nothing about that offer is unusual. What turns it into churn is surprise. Print the renewal price in dollars next to the first-order price on the product page, in the cart, at checkout, in the order confirmation and in the upcoming order email. Recharge sends upcoming order notifications three days before the charge by default (Recharge Help Center). Check that yours names the amount, and if it runs through Klaviyo, audit it with your other flows (see how much revenue should come from email). Our guide on subscribe and save discounts covers how to set the two discounts.

Where the law stands in October 2026

The FTC's 2024 "click-to-cancel" amendments to its Negative Option Rule never took full effect. The Eighth Circuit vacated them on July 8, 2025, days before full enforcement was due on July 14 (Latham & Watkins), and the FTC restored the rule's earlier text effective February 12, 2026 (Federal Register). On March 11, 2026, the FTC opened a new rulemaking by asking for public comment on whether and how to change the rule (FTC). As of October 4, 2026, the FTC's rule page lists no proposed rule after that notice (FTC).

That does not leave subscriptions unregulated. The FTC summarizes the Restore Online Shoppers' Confidence Act (ROSCA) as requiring sellers to clearly disclose all material terms before taking billing information, get express informed consent before charging, and offer simple ways to stop recurring charges (FTC). State automatic renewal laws add their own rules. California's requires consent to the offer terms, including terms offered at a promotional or discounted price for a limited time, online cancellation for subscriptions bought online, and notice 7 to 30 days before a fee change on an existing subscription (California BPC 17602). This is a summary, not legal advice. Have counsel review your checkout, emails and cancel flow.

How we audit subscription churn at Succession

When a founder brings us churn they cannot explain, we do not start with a new cancel flow or a win-back sequence. We start by finding out where subscribers actually leave.

1. One market first. We isolate US subscribers before reading any rate, because mixed markets blend different laws, payment methods and delivery times.
2. Who processes cancels. We split cancellations by channel, self-serve flow versus support, chat, phone and any bot, and measure what share of leavers ever saw a save offer.
3. A save offer in every channel. We check that support macros, chat, SMS and any AI agent offer the same skip, delay, swap, frequency and pause options as the portal, and still cancel promptly when asked.
4. A reason on every cancel. We check that manual cancels carry a reason from the same list as the flow, then read reasons by subscriber age and by product.
5. Failed payments and hard declines. We separate failed-payment cancels from customer decisions, check that hard declines get a card-update message, and review the retry and recovery setup.
6. Subscriptions with no next charge date. We count active subscriptions that will never bill, grouped by the error that stranded them.
7. Cohort payback. We read each acquisition month on net revenue and real gross margin, with Shopify orders as the scoreboard, so a save offer that keeps a subscriber at a loss does not pass for a win.

The data pulls are scripted on our side. The questions are not secret: any founder can ask their team for these answers this week. For the economics behind the last step, see our guide on LTV to CAC for subscription supplement brands.

What should you do this week?

1. Export the last 90 days of cancellations with the reason, the channel and the subscriber's age (first renewal, second, third, later).
2. Split them into portal cancels, support cancels, failed-payment cancels and stuck subscriptions with no next charge date.
3. Write one support macro per top cancel reason, with the matching skip, delay, swap or frequency offer and a prompt cancel if the customer repeats the request.
4. Make a cancel reason required on every manual cancel, using the same list as your self-serve flow.
5. Turn on a card-update message for hard declines and check that it links straight to a payment update page.
6. Print the renewal price in dollars on the product page, at checkout, in the confirmation and in the upcoming order email.
7. If you would rather run the split with someone, book a free growth audit and we will walk through your cancels, failed payments and stuck subscriptions with you, live.

FAQ