How to start a subscription box: the arithmetic first

Most subscription box plans fall apart on one line of a spreadsheet, and it is not the line people check. It is not the cost of the box. It is the number of subscribers who cancel in month two.

This post is arithmetic rather than inspiration. Every figure in it is illustrative — made up to show the shape of the calculation — so replace each one with your own and redo the sums. If the result still works, you have a business worth building. If it does not, you have saved yourself a warehouse full of cardboard.

Start with contribution per box

Contribution is what one box leaves behind after everything that box itself costs. Not profit. Just the money available to cover your fixed costs.

That is a contribution margin of about 31%. Note what is already missing: no acquisition cost, no storage, no software, no tax, and nothing for you. Those come next, and this $10.78 is the only thing paying for them.

The single most common error at this stage is costing the goods and forgetting the box. Packaging plus shipping plus pack labor is $10.50 here, which is more than the products inside. For physical subscriptions that ratio is normal, and it is why "add a nicer box" is an expensive instinct.

Two break-even numbers, not one

Add up what you pay every month whether you ship ten boxes or two hundred.

Break-even subscribers equal fixed costs divided by contribution: $319 divided by $10.78 is 29.6, so 30 subscribers. That number is where most plans stop, and it feels encouraging.

Now pay yourself $1,500 a month, which is not a salary, it is barely a part-time wage. Fixed costs become $1,819, and break-even becomes 169 subscribers. That is the real number, and it is five and a half times the comfortable one.

The platform line above is the $79 Grow plan from our pricing, chosen because a subscription business needs a custom domain and email campaigns from day one. Launch at $29 would work for a test, and the fourteen-day trial needs no card.

Shipping is the line that kills boxes

Carriers price in tiers. Cross a weight or dimension threshold by a quarter of an inch and the rate steps up, and it steps up on every box forever.

Take shipping from $6.00 to $9.50 — one size tier, or one heavier item added in month four — and contribution falls from $10.78 to $7.28. That is a 32% cut in the only number that pays your bills, and break-even moves from 30 subscribers to 44. Paying yourself now needs 250.

So design the box to the shipping tier rather than designing the box and then discovering the tier. Weigh a mock-up with real filler before you order a thousand of anything, and get a real quoted rate for the postcodes you expect, not an average.

Churn decides whether this is a business

As a rough planning simplification, average subscriber lifetime in months is one divided by your monthly churn rate. At $10.78 contribution per box:

Now put acquisition against it. Suppose a subscriber costs $18 to acquire, all in. At 10% churn you get $107.80 back, which is a healthy ratio. At 30% churn you get $35.57, and once you count the hour of support and the replacement box for the parcel the carrier crushed, you are working for nothing.

Payback matters as much as the ratio. At $18 acquisition cost and $10.78 contribution, you are repaid partway through the second box. A subscriber who cancels after one box has handed you $10.78 against $18 spent, and you shipped them something nice to achieve that.

One warning about measuring this: blended churn lies. First-month churn is always the highest, so an average across your whole base hides it. Track cohorts — the people who joined in March, followed month by month.

Why month three is where most boxes die

Three things arrive at once.

The novelty cohort leaves. A meaningful share of sign-ups wanted the first box, not a subscription. They are gone by month two or three regardless of how good the curation was.

The cash has been spent. You paid acquisition in month zero and bought goods for three boxes since. Contribution arrives $10.78 at a time while inventory is paid for up front, so a growing box is cash-hungry exactly when it looks successful.

Curation fatigue. By month three you have shipped your three best ideas. The fourth box is due in two weeks, and so is the fifth, every month, forever. This is the part nobody budgets for, and it is the reason founders quit boxes that are technically profitable.

A cohort with front-loaded churn — illustrative at 22% in month one, 15% in month two, then 8% a month — looks like this:

Everything you do about churn should concentrate on months one and two, because that is where two thirds of your eventual losses happen.

The growth equation nobody writes down

Your subscriber count does not grow forever on a fixed number of sign-ups. It settles. Steady-state subscribers equal monthly gross additions divided by monthly churn rate.

Twenty new subscribers a month at 10% churn settles at 200 subscribers, and no amount of patience moves it past that. To hold the 169 you need in order to pay yourself, you need about 17 new subscribers every single month, permanently. At 20% churn you need 34 — churn does not just shrink lifetime value, it doubles your marketing bill.

What you can actually do about each line

Price. Offer a prepaid three or six-month plan. It takes cash in up front, which solves the inventory squeeze, and prepaid subscribers structurally cannot churn mid-term.

Goods. Fewer, better items beat more items. Subscribers remember one thing they loved and forget five fillers, and the fillers are what pushed you into the next shipping tier.

Shipping. Design to the tier. Consider a flat mailer instead of a box where the product allows it, and compare regional carriers against the national one for your densest postcodes.

Churn. The first box experience carries disproportionate weight, so spend there. Offer skip-a-month instead of cancel. Tell subscribers what is coming next, because anticipation is the cheapest retention tool there is.

Failed payments. Some of what looks like churn is an expired card. Retrying failed payments on a schedule and emailing the customer recovers subscribers who never intended to leave, and this is pure found money.

Acquisition. A box is the most giftable product in commerce. Gift subscriptions and referral credit cost less than ads and arrive pre-endorsed.

What software handles, and what it will not

Recurring billing, retrying failed cards, the support inbox, scheduled campaigns and the monthly "what is in this box" announcement are all mechanical, and automating them is the difference between a box you run and a box that runs you. That is the part mAutomate handles. Our subscription box page covers how that side works, and the AI drafts rather than publishes — nothing goes out to your subscribers until you approve it.

What no platform does is source your products, negotiate your freight rate, or make box four interesting. Those are the whole job, and they are still yours. If you are still choosing between recurring and one-off selling, how to price a product covers the margin side, and how to start an online store covers the setup sequence. When you are ready to put real numbers behind this, get started.