Subscription box unit economics live or die in month two. Industry data shows roughly one-third of subscribers cancel inside the first three months, and for meal kits the figure runs 60-70%+ — yet most operators model lifetime value on the steadier churn that comes after month six. The result is a contribution margin gap that hides behind a growing top line until the second-month cohort cliff makes payback math impossible to ignore.

Why Subscription Box Unit Economics Break in Month Two

Every subscription box business has the same first-month problem: novelty masks reality. A new subscriber pays full price (or a discounted intro), receives the first box, and posts about it. Churn looks low because the cancel button has not been pressed yet — the option to cancel by month two has not even matured in the customer's mind.

Then month two ships. The customer now has a stack of last month's product, a credit card charge that wasn't anticipated, and a curiosity that has already been satisfied. McKinsey's foundational subscription research found that more than one-third of subscribers cancel within three months and more than half within six, and that cancellation rates are similar across replenishment, curation, and access models — meaning no business model is structurally immune.

The damage is not the cancellation itself. The damage is what the cancellation does to the LTV number you used to justify your customer acquisition cost. If you assumed eight months of average tenure and you actually got 2.4, every dollar of CAC you spent was sized for a customer who does not exist.

The Practical Takeaway

Stop reporting "monthly churn" as a single number. Report it as a curve: month-1 churn, month-2 churn, month-3 churn, and steady-state churn after month six. If you only have one number, you are averaging a cliff with a plateau and getting a slope that is dangerously wrong.

The Real Numbers: What Public Subscription Boxes Disclose

The cleanest public datasets come from meal-kit operators that filed under SEC reporting requirements and from analysts who reconstructed their cohorts.

  • Blue Apron: Wharton's Daniel McCarthy reconstructed Blue Apron's cohort data and found that nearly 70% of customers churned before the 4.5-month break-even point. Roughly 70% of recent Blue Apron customers were projected to never recover their acquisition cost.
  • HelloFresh: McCarthy's cohort analysis of HelloFresh showed approximately 83% of U.S. customers churned within six months. The January 2022 cohort had roughly 9% remaining by month eleven. A Bernstein note characterized the business as built on average discounts above 20% combined with 90% of customers not purchasing in Q4.
  • Dollar Shave Club: Held a comparatively benign retention profile — Retail Dive's reporting after the Unilever acquisition cited average subscriber tenure around 20 months. The replenishment-essential nature of razors flattens the second-month cliff that curation and meal kits suffer from.

The Recurly Research benchmark for 2025 puts consumer goods and retail subscription monthly churn at roughly 4.1% — 3.3% voluntary and 0.8% involuntary. But Recurly notes that subscription boxes specifically see involuntary churn reaching up to 30% of total churn in some segments, and across the broader industry estimated $129 billion in 2025 revenue loss attributable to failed payments.

The Practical Takeaway

If your blended monthly churn is anywhere north of 8-10% and you are not in replenishment, you are inside the meal-kit failure zone. Model the curve before you scale paid acquisition — not after.

How Second-Month Churn Destroys Contribution Margin

Contribution margin per box is rarely the problem in subscription commerce. The problem is how many boxes a subscriber actually receives.

Walk through the math on a curation box with these typical assumptions:

  • Retail price per box: $35
  • COGS (product + fulfillment + shipping): $20
  • Contribution margin per box: $15
  • CAC (paid social + influencer + intro discount): $45
  • Required boxes to break even: 3

If month-2 churn is 35% and month-3 churn is another 20% of the remainder, the cohort math looks like this:

  • 100 customers receive box 1 — $1,500 contribution
  • 65 receive box 2 — $975
  • 52 receive box 3 — $780
  • Cumulative contribution after 3 boxes: $3,255
  • CAC paid for 100 customers at $45: $4,500
  • Cohort is $1,245 underwater entering month four.

The cohort does eventually break even — but only if month-4-plus churn drops below 10% and the retained tail is long. The McKinsey research above suggests that for non-replenishment categories, the tail is not long enough to recover the front-loaded loss.

The Practical Takeaway

Build a simple Excel template that lays out monthly cohort survival, contribution margin per surviving customer, and cumulative recovery against CAC. If your model only computes blended LTV/CAC, replace it. The cohort view is the only one that exposes the second-month problem before it becomes a balance-sheet problem.

The Five Mechanics Driving Second-Month Cancellation

Across the McKinsey research, Recurly benchmarks, and the public meal-kit cohort data, five drivers explain most of the second-month damage:

  1. Novelty decay. Curation boxes (beauty, snacks, hobby) lose their wow factor once the customer has seen the format. The second box feels predictable even when the contents are not.
  2. Accumulation. Physical product subscribers run out of shelf space and ingestion capacity. Meal kits churn because the customer hasn't eaten last week's box. Beauty boxes churn because the bathroom counter is full.
  3. Intro-discount cliff. Customers who joined at 50% off see the full price on the second invoice. Bernstein's HelloFresh note specifically called out average discounts above 20% as a structural retention liability.
  4. Involuntary churn. Recurly's data shows failed payments cause a meaningful slice of cancellations that the customer never actively chose. Card expiry, fraud declines, and address mismatches compound in month two when the auto-renewal is no longer "the first charge."
  5. Effort tax. If the box requires the customer to do something — cook the meal, assemble the craft, fit the clothing — the second box exposes whether they actually want the work. The first box was a vote of intent; the second is a vote of habit.

The Practical Takeaway

Diagnose which of the five is dominant in your category before you build the retention playbook. Discount-cliff and involuntary-churn solutions look nothing like novelty-decay solutions, and most operators try the wrong fix because they never tagged the cancellation reason.

A Step-by-Step Framework for Modeling Month-Two Risk

This is the spreadsheet model every subscription box operator should run weekly. Most don't — they run a P&L instead, which hides cohort behavior under aggregate revenue.

  1. Pull cohort data by signup month. Every customer's first-box month is their cohort. Track how many of that cohort received box 2, box 3, and so on, by elapsed month.
  2. Compute survival curves, not churn rates. Survival is the cumulative percentage of the cohort still receiving boxes at month N. Plot it. The shape matters more than any single number.
  3. Compute contribution margin per cohort per month. Multiply surviving subscribers by per-box contribution. Add discounts and refunds as a negative line.
  4. Compute cumulative contribution against CAC. Each cohort has a payback month — the elapsed month at which cumulative contribution crosses cumulative CAC. If that number is greater than six, you have a structural problem.
  5. Run the same cohort against three scenarios. Hold month-1 churn flat, cut month-2 churn by 25%, and cut month-2 churn by 50%. Look at how much the payback month shifts. Almost always, month-2 levers move payback more than acquisition-side levers do.
  6. Set a kill criterion. Decide in advance: if a paid channel's cohort payback exceeds X months, you turn off spend on that channel. Most operators never define this and end up scaling losses.

The Practical Takeaway

The spreadsheet model is the operating system of a subscription business. Without it, you are flying on revenue growth and gut feel — exactly the conditions under which Blue Apron, Birchbox, and every other casualty of the 2010s subscription boom lost the plot.

Tactical Retention Moves That Reset Month-Two Churn

Operators with healthy cohorts share a pattern of small interventions specifically targeted at the month-1-to-month-2 transition. None of them are revolutionary. All of them are deliberate.

  • Pre-month-two value injection. Send a high-leverage benefit between box 1 and box 2 — a personalization onboarding flow, a content unlock, a referral credit — so the second charge feels like an upgrade rather than a renewal.
  • Skip-a-month, not cancel. Friction matters, but ethical friction does not. Offering a one-month skip captures customers who would otherwise cancel and gives them a graceful re-entry path.
  • Card-update flows. Recurly's $129B involuntary churn figure exists because most operators do not have a serious dunning system. Pre-expiry email plus in-app re-auth plus retry sequencing typically recovers 35-50% of failed payments.
  • Right-size box cadence. If your customers accumulate inventory, offer bi-monthly. Stitch Fix's model — schedule-on-demand rather than fixed monthly — exists because the original fixed cadence created accumulation churn.
  • Anchor the discount. If you must discount to acquire, make the intro discount span two or three boxes instead of one. This pushes the full-price cliff to a point where habit has formed.

The Practical Takeaway

Pick two of the five and instrument them this quarter. A 25% improvement in month-2 retention typically produces a larger LTV gain than a 25% CAC reduction, and it is cheaper to engineer.

Conclusion: Build the Model Before You Scale the Spend

The pattern across two decades of subscription commerce is consistent. Operators who survived the second-month cliff (Dollar Shave Club, Netflix, Spotify, replenishment-essential boxes) did so because their category structurally avoided it, or because they modeled cohort survival from day one and shut off acquisition channels that did not pay back. Operators who did not (Blue Apron pre-acquisition, Birchbox pre-acquisition, the long tail of curation boxes that quietly closed) built their plans on blended churn and were surprised by the math that was always there.

You do not need a data team to do this. You need a spreadsheet with cohort survival rows, contribution margin columns, and a CAC payback formula. A pre-built subscription unit economics template — the kind operators in this category should be running before they spend the next $10K on paid social — turns the abstract second-month risk into a number you can see, debate, and act on. That is the entire job.

Sources

Related: Browse all Best Financial Model Templates on ModelStack.

Get started with a free template

Download our free Unit Economics Calculator — no signup required.

Download Free Template