The Shopify Subscription Business Model: Unit Economics, Churn Math, and When It Actually Works

Most Shopify brands add subscriptions because a podcast told them recurring revenue multiplies their exit valuation. Then they wire up Recharge, slap a 15% “Subscribe & Save” badge on the PDP, and watch 40% of their one-time buyers migrate to the discounted SKU without adding a single new customer.

That’s not a subscription business. That’s a 15% price cut with extra app fees and a support queue full of “cancel my order” emails.

The subscription model works — but it works on specific math, for specific product categories, at specific churn rates. If your month-3 retention is below 50% and your discount is 20%, you are paying customers to leave slightly slower. This article is the operator version: the unit economics, the churn thresholds, the cannibalization test, and the decision framework for whether to build it at all.

The Only Equation That Matters

Forget LTV:CAC ratios for a second. The question that decides whether your subscription program is accretive is simpler:

Does the gross profit from the average subscriber’s full lifetime exceed the gross profit from that same person as a repeat one-time buyer, net of discount and app costs?

Run it concretely. Say you sell a $48 coffee bag with $14 COGS and $7 shipping/fulfillment per order.

One-time buyer: $48 revenue − $14 COGS − $7 fulfillment − $2.00 payment fees = $25.00 gross profit per order. They reorder 2.4 times a year at a 38% repeat rate. Call it $60/year in GP from the cohort average.
Subscriber at 15% off: $40.80 revenue − $14 COGS − $7 fulfillment − $1.78 fees = $18.02 gross profit per order.

So the subscriber needs to complete 3.4 orders just to match the one-time buyer’s annual GP. If your average subscription life is 2.8 orders — which is dead normal for consumables in year one — you have made the customer less valuable by converting them.

This is the trap. Subscriptions only win when order frequency increases materially, not when you discount the same frequency.

The frequency lift test

Before you launch, pull your repeat purchase interval from Shopify’s order data. If your best customers already reorder every 35 days and your subscription cadence is 30 days, you’re buying a 14% frequency lift for a 15% discount. Net negative.

If your average reorder interval is 90 days on a product that’s physically consumed in 35 days, you have a real gap. Subscription closes it. That’s a 2.5x frequency lift for 15% off — that’s the trade worth making.

Churn Math: The Numbers You Need to Hit

Subscription churn in DTC is brutal and front-loaded. Benchmarks from consumables brands in the $2M–$20M range:

Order 1 → Order 2 churn: 25–40%. This is the single biggest leak. Most of it is “I forgot I subscribed” or “I don’t need it yet.”
Order 2 → Order 3 churn: 15–22%.
Steady-state monthly churn after order 4: 5–9% for well-run programs, 12%+ for weak ones.

Average subscription lifetime at 8% monthly churn ≈ 12.5 cycles. At 12% monthly churn ≈ 8.3 cycles. At 18% ≈ 5.5 cycles. That difference — 8% vs 18% churn — is the difference between a $225 LTV and a $99 LTV on the same product.

The three churn numbers to put on a dashboard

Active subscriber count (not “total subscriptions created” — vanity).
Order-2 completion rate. Target 70%+. Below 60% means your onboarding or cadence is broken.
Net MRR movement: new subscriber MRR + reactivation + upgrades − churn − downgrades − skips. Skips matter. A 20% skip rate on any given cycle is a 20% revenue haircut even if nobody cancels.

Skips are the silent killer in Shopify subscription programs. Operators track cancellation religiously and ignore skips entirely. If 1,200 active subscribers skip at 18% in a month, that’s 216 orders × $40.80 = $8,800 in revenue you forecast and didn’t get.

Cannibalization: Measuring the Real Cost

Here’s the test most brands skip. Thirty days after launch, segment your subscription signups:

Net-new customers who bought subscription on first order.
Converted repeat buyers who were already reordering at a healthy clip.

Group 2 is cannibalization. If your subscriber base is 65% converted repeat buyers and your discount is 15%, calculate the drag:

650 of 1,000 subscribers × ~4 orders/year × $7.20 discount = $18,720/year in discount given to people who were buying anyway.

That’s only acceptable if those 650 people now order 5.5 times instead of 4 — a 37% frequency lift. Check it against their pre-subscription interval. Most brands never do, which is why subscription programs frequently show flat net revenue with a lower blended margin.

How to reduce cannibalization

Replace the flat discount with a value ladder. 10% off order 1, 15% at order 3, 20% at order 6. You pay for retention, not for signup.
Gate the deepest discount behind prepaid plans. A 3-month prepay at 20% off gives you cash upfront, eliminates orders 2–3 churn entirely, and shifts payment risk forward.
Offer non-discount value. Free shipping (if you charge it), a subscriber-only SKU, early access, or a free add-on product with 80%+ margin. A $6 COGS gift costs you less than a $7.20 discount and feels like more.

The Hidden Cost Stack

Subscription revenue carries costs that one-time revenue doesn’t. Budget for all of them:

Subscription app: Recharge, Skio, Loop, Stay AI. Expect $99–$499/month base plus 0.5%–1.0% of subscription revenue plus $0.19 per transaction on some plans. On $80K/month in subscription GMV, that’s $600–$1,100/month all-in. Shopify’s native Subscriptions app is free but thin on dunning and retention tooling.
Failed payments (dunning). 5–12% of recurring charges fail on first attempt — expired cards, insufficient funds, issuer declines. A good dunning sequence recovers 55–70% of those. A bad one recovers 25%. On $80K/month, the difference between good and bad dunning is roughly $2,900/month in recovered revenue.
Support load. Subscribers generate 2–3x the ticket volume per customer of one-time buyers. Budget $0.80–$1.50 per subscription order in support cost at typical ticket rates.
Chargebacks. Subscription chargebacks run 2–4x higher than one-time. “I didn’t authorize this” is the most common reason code. Every one costs $15–$25 in fees plus the lost goods.
Fulfillment complexity. Subscription orders batch on specific days, which creates pick-pack spikes. If you’re 3PL-based, ask whether your peak-day volume triggers surge pricing.

Add it up and the true cost of subscription infrastructure is typically 3–5% of subscription revenue on top of your discount. Model it that way or your contribution margin reporting is fiction.

Which Products Actually Work

Subscription fit is mostly determined by the product, not the marketing.

Strong fit

Consumables with a predictable depletion rate: supplements, coffee, pet food, skincare, cleaning products, razor blades. The customer runs out on a schedule. You’re solving a real forgetting problem.
Replenishment cycle of 30–60 days. Shorter than 30 and you get stockpiling complaints. Longer than 90 and the customer forgets they subscribed and churns on sight of the charge.
AOV of $35+. Below that, fulfillment and payment fees eat the discounted margin alive.
Gross margin above 60% pre-discount. Below 55%, a 15% subscription discount takes you under the threshold where CAC payback works.

Weak fit

Apparel, home goods, accessories. No depletion. Curation-box models can work but churn is 2x consumables and inventory risk is severe.
Products with long, irregular reorder intervals (6+ months). You’ll get signups and almost no second orders.
Anything where variety matters more than consistency. If your repeat buyers rotate SKUs every order, a locked-in subscription fights their behavior.

The Retention Levers That Actually Move Numbers

Ranked by impact per hour of work:

1. Pre-billing notification (biggest single lever)

Email 3 days before charge, with a one-click “skip this one” and “change date” link. Counterintuitively, making it easy to skip reduces cancellations. Brands that add a skip option typically see cancellation drop 20–30% while skip rate rises 10–15% — a net positive because a skipper stays in the file and a canceller doesn’t.

2. Fix order-2 churn specifically

A 10-point improvement in order-2 completion (from 62% to 72%) lifts total program LTV by roughly 16% because every downstream cohort gets bigger. Tactics: a day-7 “how to use it” email, a day-21 “you’re about to run low” email, and a cadence adjustment prompt if the first delivery arrived too soon.

3. Smart dunning

Retry on day 1, 3, 5, 7, 12 — not five consecutive days. Pair with SMS on attempt 2 and an easy card-update link. Moving from 35% to 60% recovery on a $80K/month program is worth roughly $25K–$30K annually in pure margin.

4. Cancellation flow with offers

Before the cancel button, offer: skip one, delay 30 days, swap product, or reduce quantity. A well-built flow deflects 20–35% of cancellation attempts. Most of those deflections convert into 2+ additional orders.

5. Prepaid upgrade offers

At order 3, offer a 6-month prepay at 22% off. Customers who accept effectively have 0% churn for the term, and you collect roughly $245 in cash today instead of $40.80/month with 8% monthly leak. Even at a deeper discount, prepaid cohorts usually show 25–40% higher realized GP per customer.

Modeling It Before You Build It

Build a simple cohort sheet in a spreadsheet before you install an app. Columns: cycle number (1–18), surviving subscribers, revenue per order, GP per order, cumulative GP. Apply your churn curve — not a flat rate, a front-loaded one (35%, 20%, 12%, 9%, 8%, 8%…).

Then answer three questions:

What’s cumulative GP per subscriber at cycle 12? Compare to your blended one-time customer’s 12-month GP.
At what cycle does a subscriber break even against CAC? If you’re paying $42 CAC and the subscriber generates $18 GP/order, you’re profitable at order 3. If your order-3 completion is 48%, half your acquisition is unprofitable.
What churn rate makes this break even? Solve for it. If the answer is “anything under 9% monthly,” and industry norm for your category is 11%, don’t launch until you’ve built the retention stack first.

Subscription programs fail quietly. Revenue looks fine on the dashboard, margin drifts down 3–4 points over six months, and nobody connects the two because the discount lives in a different report than the COGS. Track subscription contribution margin as its own line item from day one.

A Realistic Launch Sequence

Weeks 1–2: Pull reorder interval data by SKU. Identify the 2–4 products with a real depletion gap. Model the cohort sheet.
Weeks 3–4: Install the app. Launch on one hero SKU only, at 10% off, with a prepay option at 18%. Don’t badge the whole catalog.
Weeks 5–8: Build the pre-billing email, the dunning sequence, and the cancellation flow before you push volume. These are non-optional infrastructure, not optimizations.
Month 3: Measure order-2 completion. If it’s under 60%, fix cadence and onboarding before expanding.
Month 4–6: Expand to additional SKUs. Introduce the loyalty ladder. Test prepaid upgrade at order 3.
Month 6: Run the cannibalization analysis. Compare blended gross margin to pre-launch baseline. If it dropped more than 1.5 points without a matching revenue lift, cut the discount depth.

The brands that make subscriptions work treat it as an operations problem — churn curves, dunning recovery, cadence fit — not a marketing badge. The discount is the least interesting part of the model and the part everyone over-indexes on.

Try ProfitAnalyzer free — no credit card required

Find your Shopify margin leaks in under 2 minutes. Now available on the Shopify App Store.

Install on Shopify →Try free online →

The Shopify Subscription Business Model: Unit Economics, Churn Math, and When It Actually Works