Healthy supplement subscription retention runs 75-85% of a cohort still active after the first renewal, 65-75% at month two, 58-68% at month three, 42-52% at month six, and 28-38% at month twelve. Below those ranges the program is leaking. A single blended churn rate is close to useless in this category, because retention behaves differently by month, by acquisition source, and by product form, and blending averages all of that into a number nobody can act on.
Every benchmark on this page is my own operator range, taken from health and wellness DTC brands between $1M and $60M+, not from a published study. I’ve read the subscription file inside brands at most of those stages, including the leadership rebuild that carried Ancestral Supplements from $28M to $60M+ and the $6M practice Dr. Gabrielle Lyon runs with recurring revenue underneath it. The curve below is what I expect to see before I look, and the gap between it and what I find is the diagnosis.
The month-by-month retention curve
Retention here means the percentage of subscribers from a single acquisition cohort still active in month N. Month 0 is the first order. Month 1 is measured after the first renewal charge clears, which is the number most brands get wrong because they count subscribers who cancelled before ever being billed twice.
| Month | Healthy | Warning | Broken |
|---|---|---|---|
| Month 1 | 75-85% | 65-75% | Under 65% |
| Month 2 | 65-75% | 55-65% | Under 55% |
| Month 3 | 58-68% | 45-58% | Under 45% |
| Month 6 | 42-52% | 30-42% | Under 30% |
| Month 12 | 28-38% | 18-28% | Under 18% |
Read the shape of the curve before you argue with any individual cell. A brand at 80% in month 1 and 48% in month 3 has a steeper problem than a brand at 70% and 60%, even though the first one looks better at the top. What you want is a curve that flattens. Once a subscriber clears month 4 or 5, monthly churn in supplements usually settles into the 4-8% range and stays there, and that flat stretch is where the compounding value lives.
Why month two is the cliff
The biggest single drop in supplement subscriptions happens between the first renewal and the second. In the brands I’ve looked at, the month 1 to month 2 window accounts for roughly 30-45% of all churn a cohort will ever produce. Everything after month 4 combined is usually smaller.
Month 2 is the first charge the customer never consciously agreed to. Month 0 was a purchase decision. Month 1 often arrives while the customer still feels good about the brand and hasn’t finished the first bottle. By month 2 a second unopened bottle is sitting next to the first, the card gets charged for product they haven’t consumed, and the value question turns concrete. Fix that window and the entire curve shifts up, because every later month is calculated off a larger surviving base.
The formulas
Definitions matter more than the arithmetic here. Two brands can report churn rates 8 points apart while running identical businesses, purely because one counts failed payments as churn and the other doesn’t.
Formula: Monthly subscriber churn = subscribers lost during the month ÷ active subscribers at the start of the month
Formula: Monthly retention rate = 1 − monthly subscriber churn
Formula: Average subscription lifetime in cycles = 1 ÷ monthly subscriber churn
Formula: Subscription LTV = contribution margin per subscription order × average subscription lifetime in cycles
Formula: Subscription CAC payback = blended acquisition cost ÷ contribution margin per subscription cycle
Use contribution margin in the LTV calculation. Revenue-based subscription LTV is the most commonly inflated number in DTC, and it’s usually the one being compared against a CAC figure measured in real dollars today. The full set of numbers this sits inside is in the ecommerce metrics that predict scale.
One caution on the lifetime formula. Dividing 1 by churn assumes a constant monthly churn rate, and supplement churn is front-loaded, so the formula overstates lifetime for young programs. With under 12 months of history, use actual cumulative margin at a fixed horizon and treat the formula as a sanity check.
Why the blended number lies
Your reported churn rate is a weighted average of populations that behave nothing like each other. Four splits change the answer more than anything else.
Acquisition channel. A subscriber acquired through discount-led paid social churns materially faster than one who came from organic search, email, or a referral. I regularly see a 15-25 point spread in month 3 retention between the best and worst channels inside the same brand. If your paid mix shifted last quarter, your blended churn moved for reasons that have nothing to do with your retention work.
First-order offer. Deep first-order discounts buy a worse cohort. A 50% off first-box offer produces subscribers whose month 2 retention often runs 10-20 points below the same brand’s 15% off cohort, because a real share of them subscribed to get the discount and planned to cancel from the start. The offer selects your customer before it converts them.
Product form and dose duration. A 30-day supply on a 30-day cadence has almost no slack. A 60-count bottle at two capsules a day, shipped every 30 days, quietly stacks inventory in the customer’s cupboard and produces a cancellation in month 3 or 4 that looks like dissatisfaction and is actually arithmetic. Powders and tinctures with variable dosing churn differently again, because the real consumption rate is unknown to both of you.
Chosen versus defaulted. Subscribers who actively selected subscribe-and-save retain far better than subscribers opted in by a pre-checked box or a price display that made one-time purchase look punitive. Defaulted subscribers inflate your subscriber count on day one, then show up as month 1 and month 2 churn plus a chargeback rate you didn’t want.
How to read the cohort report
Build it once and it answers most retention questions permanently. Rows are acquisition month. Columns are months since first order. Each cell is the percentage of that cohort still active. Twelve months back, minimum. One export and a pivot table gets you there.
Fix the horizon before you look. Month 3 is the decision horizon, fast enough to act on and predictive enough to trust. Month 12 is the valuation horizon, since that number sets what your recurring revenue is worth to a buyer. Pick both and never quote a retention figure without naming which one you mean.
The comparison that matters is one cohort against another at the same month count. January’s cohort at month 3 against April’s cohort at month 3. That’s the only version that isolates whether the customers you’re buying are getting better or worse, because it holds maturity constant. Comparing a young cohort’s month 1 to a mature cohort’s month 8 is how brands talk themselves into believing retention improved when their acquisition mix simply got younger.
Then split the same table by acquisition channel and by first-order offer. If you build only one additional view, build the channel split. It changes media buying decisions, which is where the money is.
The four causes of month-two churn, and the fix for each
Cadence mismatch. They still have product. Match the ship interval to actual dose duration per SKU instead of defaulting everything to 30 days, and send a pre-billing notification that makes delaying a shipment easy. Brands that let customers set 45 and 60-day intervals often watch total revenue per subscriber rise even as shipment frequency falls.
Value gap. They don’t feel anything yet, and nobody told them when they should. Most supplements need 4-12 weeks before an effect is noticeable. If your onboarding doesn’t state that timeline explicitly in the first two weeks, the customer invents their own expectation and cancels when it isn’t met. Cheapest fix on this list, and the one most brands skip.
Friction. They can’t easily skip, so they cancel instead. Every barrier between a customer and a skip button converts a pause into a permanent loss. Put skip and delay in the account portal and in the pre-billing email, one click each. Retention programs that only fire discounts at the cancel page are treating a self-service problem with a margin giveaway. What holds up over a longer horizon is covered in customer loyalty campaigns for supplement brands.
Payment failure. Involuntary churn from expired cards, insufficient funds, issuer declines, and bank fraud holds is frequently 20-40% of total subscription churn in the brands I audit. Most founders have no idea, because it never appears as a cancellation reason. A dunning sequence with 4-5 retry attempts spread across 21 days, card-updater services enabled through your processor, and an email that treats the failure as a payment problem rather than a sales pitch will typically recover 30-50% of failed charges. That’s an ops fix with a marketing-sized return, and it’s usually the fastest retention win available.
What good looks like at each stage
| Revenue | Subscriber share of revenue | Month 3 retention | Fix this first |
|---|---|---|---|
| $1M | 15-25% | 50-60% | The first-order offer and the month 1 to 2 drop |
| $5M | 30-45% | 55-65% | Involuntary churn and the channel-split cohort report |
| $20M+ | 45-60% | 60-70% | Cadence by SKU and retention-weighted media buying |
At $1M, stop optimizing the cancel page and look hard at what your first-order offer is buying you. A $5M brand gets more out of a dunning sequence and a working cohort report than out of any new lifecycle flow. Past $20M the money moves when retention-by-channel data feeds back into acquisition, so you stop paying full price for subscribers who die at month 2.
When chasing retention is the wrong priority
Two situations make retention work a wasted quarter.
The first is a product efficacy problem. If month 3 retention sits under 40% across every channel and every offer, and your reviews and support tickets say people don’t feel anything, no email sequence fixes that. Formulation and dosing come first, along with what you tell people to expect and when. Retention tactics applied to a product people don’t want make the churn slower and more expensive.
The second is a broken first-order offer. If you’re acquiring subscribers at a cost that requires 8 cycles to pay back and your median subscriber makes it 3, you’re working the wrong end of the funnel. Change the offer or change the price. When acquisition looks fine and retention is the real constraint, that diagnostic runs differently.
One more thing worth naming: churned subscribers aren’t gone. A well-built win-back sequence for lapsed subscribers reactivates a meaningful share of month 2 and month 3 cancellations at a fraction of new-customer acquisition cost.
Common questions
What is a good subscription retention rate for supplements? 75-85% after the first renewal, 58-68% at month three, 42-52% at month six, and 28-38% at month twelve, measured by acquisition cohort. Under 45% at month three, something structural is wrong with either the offer or the product itself.
What percent of subscribers churn in the first month? In healthy supplement programs, 15-25% of subscribers don’t survive the first renewal. Above 25% usually means the subscription was oversold at checkout or defaulted into rather than chosen.
How much of churn is failed payments? Typically 20-40% of total subscription churn is involuntary. Most brands underestimate it because failed payments never produce a cancellation reason. Dunning plus card updater services recover 30-50% of those charges.
How do I calculate subscription LTV? Contribution margin per subscription order multiplied by average lifetime in cycles, where lifetime is 1 divided by monthly churn. Use contribution margin rather than revenue, and for programs under a year old use actual cumulative margin at month 12 instead of the formula.
Should I offer a discount to save a cancellation? Offer a skip or a longer interval first, and a discount only after those get declined. Discount-saved subscribers churn at the next renewal at a much higher rate, and you’ve permanently repriced that customer. A cadence change saves the same subscriber without giving up margin.
If you’re running a subscription program that looks fine in aggregate and you can’t tell which cohort is dragging it down, the work starts with the cohort report and the channel split. That reporting layer is what Growth OS builds first in any engagement, so the retention conversation runs on numbers we both trust.