The second charge is the first one the customer has not already decided to pay. The first order was a purchase — considered, wanted, justified. The second is a bill arriving for a decision made thirty days ago by someone who has not yet been given a reason to believe the product is working. That gap is where supplement subscriptions die, and it is why month two and month three look so different from month one.
The mistake most brands make is treating this as one number. "We're churning 40% by month three" is four separate problems averaged together, and three of them have completely different fixes. Separating them is the whole job.
Split the number before you spend anything
Involuntary churn is a payment failure: an expired card, a declined transaction, a bank's fraud logic, a customer who got a new card after a breach. The customer did not decide to leave. They are frequently unaware they left.
Voluntary churn is a decision, and it has at least three distinct causes:
- Cadence mismatch. You ship 30 days of product and they take it 20 days a month. By month three they have a shelf of unopened bottles and canceling is the rational response to a stockpile.
- The efficacy gap. They cannot tell whether it is working, nobody set an expectation for when they would, and the second charge arrives right in the middle of that uncertainty.
- Surprise and friction. They forgot they subscribed, the charge appeared without warning, and cancel was the only button they could find. Some of these become chargebacks, which is expensive twice.
The first diagnostic step costs nothing: pull the split. What fraction of your lost subscribers failed a payment versus actively canceled? Brands are frequently surprised, and the surprise usually points at the cheapest available fix. Do this before you buy a loyalty app.
Involuntary churn is an ops problem, and it is the cheapest win
This is the one place in retention where the fix is mechanical, unglamorous and reliably underdone.
Pre-dunning. Cards expire on a known date. Email before it happens, with a one-tap update link. This alone recovers subscribers who would otherwise have failed silently.
A retry schedule that reflects how banks behave. Retrying a declined card four times in twelve hours is not persistence, it is a way to get flagged. Space retries over days, vary the timing, and stop at a defined point rather than retrying forever.
Account updater. The card networks operate updater services that let a merchant receive new card details when a customer's card is reissued. Whether you have access depends on your processor and subscription platform. Ask both, explicitly, because it is often available and switched off.
Multi-channel recovery. Email, then SMS. A failed payment notice is one of the few messages a customer genuinely wants, and SMS gets read.
Say "paused", not "canceled". The state after a failed payment should be recoverable, and the language should invite a fix rather than confirm a loss. A subscriber whose card failed still wants the product; do not tell them their subscription ended.
Keep the billing descriptor recognizable. A charge from a company name the customer does not recognize generates disputes. Match the descriptor to the brand on the box.
Cadence: ship at the rate they actually consume
A 30-day supply on a 30-day renewal assumes perfect daily compliance. Nobody has perfect daily compliance. Real consumption for most supplement categories is slower than the label's serving schedule, and the stockpile that results is a visible, physical argument for canceling.
Three things fix it:
- Let the customer set frequency at checkout and change it afterwards, in one tap, without contacting support. 30, 45, 60 days.
- Make "skip this delivery" more prominent than "cancel" in the portal and in every pre-renewal email. Most people who click cancel wanted to skip. Give them the smaller action first.
- Send a pre-renewal notice with the skip button in it, timed a few days before the charge — long enough to act, short enough to still be top of mind. This is the single highest-leverage email in a supplement flow set, and many brands do not send it at all because it looks like inviting a skip. It is inviting a skip. A skip is a retained customer; a surprise charge is a cancellation and possibly a dispute.
If you sell multiple SKUs, the ability to swap products at the same subscription price keeps people who are bored rather than dissatisfied. Boredom is a much easier problem than doubt.
The efficacy gap: month two is a judgment, so give them something to judge
The customer is running an experiment on themselves with no protocol and no control group. Left alone, they will conclude nothing, and "nothing" reads as "not working".
The fix is education timed to the doubt, not to your campaign calendar:
- On delivery: how to take it, when to take it, with or without food, and the specific things they should and should not expect. Under-promising here is a retention strategy.
- Week one: what is normal so far. Reassurance beats enthusiasm.
- Week two to three: the honest expectation window for the category, stated as a range, based on what your substantiation actually supports.
- Just before the second charge: a check-in that gives them a way to notice change — a simple self-report question, a reminder of where they started. People who can point to something are dramatically less likely to cancel than people who cannot remember why they started.
- After the second delivery: the compounding message. This is where multi-month framing lands.
Every word of that has to stay inside your substantiated claims. Do not solve a retention problem by writing a claim you cannot defend — you will trade churn for a regulatory and payment-processing problem, and the label and claims guide explains why that trade is a bad one.
The cancellation flow is a product surface
Most cancellation flows are a form with a dropdown. That is a wasted opportunity and a wasted data source.
A working flow does three things:
- Asks why, with real options — too much product, too expensive, not working, side effects, just wanted to try it, switching brands. That dropdown is the most honest research you will ever collect, and it should feed a dashboard, not an inbox.
- Offers the answer to the reason given. "Too much product" gets a longer interval. "Too expensive" gets a smaller size or an annual prepay. "Not working" gets a real answer from a human, or graceful acceptance. Offering a discount to someone with a side effect is how brands earn one-star reviews.
- Keeps the door open. A pause with a return date beats a cancellation, and a clean cancellation with a good winback later beats a hostile retention wall. Making cancellation hard produces chargebacks, complaints and processor scrutiny. It is not a retention strategy, it is a deferred cost.
Prepay, and its honest downside
A three-month or six-month prepaid plan at a discount is the most direct fix for month-two churn, because the month-two decision no longer exists. Offer it at checkout and again in the cancellation flow.
The unprofitable part: prepay does not improve your product, it defers the decision. It also pulls cash forward and can hide a genuine retention problem behind healthy-looking revenue. Track prepaid cohorts separately and watch what happens at the renewal of the prepaid term — that is your real retention number. If prepay renewal is bad, the underlying problem was never billing frequency.
Measure cohorts, not a blended rate
A blended monthly churn number is nearly useless because it mixes cohorts acquired under completely different conditions.
What to track instead:
- Order-2 conversion rate by acquisition cohort. The percentage of first orders that become a second order, split by month acquired and by channel. This is the headline number.
- Involuntary versus voluntary split, tracked separately every month.
- Retention curves by acquisition channel. Discount-driven and heavily incentivized cohorts churn differently from organic ones, and averaging them hides the fact that your best-performing ad is buying customers who leave.
- Retention by SKU. Some products in your catalog retain; some do not. That should change what you promote.
- Skip rate, which should be non-zero and healthy. A skip rate of zero means nobody can find the button.
- Reactivation rate from winback flows.
- Prepaid term renewal rate, separately from monthly.
A dashboard in Metabase or Looker Studio starts at $799 and is a better first purchase than another app. You cannot fix churn you cannot see, and subscription platform reporting is usually aggregated exactly where you need it split.
What to build, in order
- Fix dunning and payment recovery. Cheapest, fastest, largest immediate return.
- Add the pre-renewal notice with a one-tap skip. Reduces surprise charges and disputes.
- Build the post-purchase education flow timed to the doubt window.
- Make skip, swap and interval changes self-service in the portal.
- Rebuild the cancellation flow with reason capture and matched offers.
- Then consider loyalty, referrals and prepay tiers.
Doing six before one is the most common sequencing error in this category, and it is expensive because loyalty apps charge monthly while dunning improvements do not.
Note that items four and five are storefront work rather than email work, which is why they stall in brands whose store was built by a generalist. The subscribe-and-save path, the cadence selector, the customer portal and the cancellation flow are the parts of a supplement store that decide repeat revenue, and we build them as the core of the store rather than as apps bolted on after launch.
What this costs
- Klaviyo email flow set-up — from $799. Welcome, abandonment, post-purchase and winback flows, built and tested.
- Retention Engine — $1,499 in 14 days. Klaviyo flows rebuilt end to end, SMS for the moments email is too slow, a loyalty program that rewards reorders, and winback sequences for lapsed buyers. Details on the supplement brands page.
- Dashboards — from $799. The cohort view above.
- Email and retention retainer — from $599/mo, if you want the flows kept tuned as the catalog changes.
Everything is on the pricing page. If your store itself is the constraint rather than the flows, the Shopify supplement store cost breakdown covers the build side.
The last honest thing: if customers cannot tell your product is doing anything and your own evidence does not support a clear expectation, no flow set fixes that. Retention systems make a good product's value visible. They cannot manufacture value that is not there, and brands that try end up writing claims that create bigger problems than churn. If you want someone to look at your actual numbers and tell you which of the four problems you have, send them over.
Frequently asked questions
Why do supplement subscribers cancel in month 2?
Usually one of four reasons: a failed payment they never noticed, a stockpile from shipping faster than they consume, uncertainty about whether the product is working, or a surprise charge they had forgotten was coming. Each has a different fix, so the first step is splitting your churn into involuntary versus voluntary and then segmenting the voluntary cancellations by stated reason.
What is involuntary churn and how do I reduce it?
It is subscribers lost to payment failures rather than decisions — expired cards, declines, reissued cards after a breach. Reduce it with pre-expiry notices, a retry schedule spaced over days rather than hours, card network account updater services through your processor, SMS as a second channel, and a "paused" state rather than an immediate cancellation.
Should I make it easy to skip a delivery?
Yes, and it should be more prominent than cancel. Most customers who cancel wanted to skip, and a skip retains the subscriber while a surprise charge risks a cancellation and a chargeback. Put the skip link in the pre-renewal email and one tap deep in the portal. A skip rate of zero is a warning sign, not a win.
Do prepaid subscription plans fix churn?
They remove the month-two decision, which shows up immediately in retention, but they defer the problem rather than solving it. Track prepaid cohorts separately and watch the renewal at the end of the prepaid term — that is the real number. Healthy-looking revenue from prepay can mask a product or expectation problem that surfaces later.
What retention metrics should a supplement brand track?
Order-2 conversion rate by acquisition cohort and channel, involuntary versus voluntary churn split, retention curves by channel and by SKU, skip rate, reactivation rate from winback, and prepaid term renewal rate. Blended monthly churn mixes cohorts acquired under different conditions and hides the thing you need to act on.
How much does it cost to fix subscription retention?
Klaviyo flow set-up starts at $799, and Retention Engine rebuilds flows, SMS, loyalty and winback for $1,499 in 14 days. A cohort dashboard is $799. Start with dunning and the pre-renewal skip email, which cost the least and return the most, before buying a loyalty app with a monthly fee attached.