Subscription payback: what you can afford per subscriber.

Ask a subscription team what they can afford to pay for a subscriber and you will often hear a benchmark, or a ratio of lifetime value to acquisition cost. Both skip the question that decides whether growth is survivable: how long until the money comes back?

That is payback. It is the month in which the cumulative contribution from a subscriber passes what it cost to acquire them. It is less glamorous than lifetime value and far more useful, because it is a cash question, and cash is what runs out.

The three inputs

• →Cost per paying subscriber. Not cost per trial. Divide cost per trial by the trial-to-paid rate.
• →Monthly contribution per subscriber. The price after app store or payment fees, refunds, and the variable cost of serving them.
• →The retention curve. The share of a starting cohort still paying at each month, from your own billing records, ideally by channel.

A worked table

Take a $15 monthly plan that leaves $12 of contribution after fees and costs, and a cost per paying subscriber of $60, from an $18 trial converting at 30%. The retention curve below is illustrative arithmetic. Use your own.

The running total passes $60 in month nine. That is the payback. By month twelve the average subscriber has brought $75, so the first year returns 1.25 times the acquisition cost. Whether that is good depends on the business. A company with a year of runway and a steady flow of new cohorts may be comfortable. A company that needs the money back inside a quarter is not.

The levers, and what each is worth

The table makes the levers concrete, and it shows why the biggest ones are rarely in the ad account.

• →Trial-to-paid rate. At the same $18 per trial, a rate of 36% instead of 30% cuts the cost per subscriber to $50, and payback moves from month nine to month seven. Six points on the trial is worth two months of cash.
• →Early retention. Most of the loss in the table happens between month one and month two. Keeping a few more people through the first renewal lifts every month after it, which moves payback more than a similar gain later in the curve.
• →Cost per trial. Cheaper trials help only if they convert at the same rate. A lower cost per trial that brings a lower trial-to-paid rate can leave payback longer than before.
• →Price and plan mix. A higher price or a larger share of annual plans raises contribution, but test the effect on trial-to-paid and retention before counting the gain.

Annual plans change the cash, not the logic

An annual plan paid upfront can recover the acquisition cost on day one. A $99 plan that leaves about $79 after fees covers a $60 subscriber immediately. That is valuable, and it is also where teams fool themselves. The cash arrived early, but the subscriber has not proven they will renew, and refunds in the first weeks come straight out of it.

Treat annual and monthly subscribers as separate cohorts. Judge annual acquisition on renewal rate at month thirteen once you have it, and until then on refunds and engagement. Mixing the two in one payback figure makes monthly acquisition look better than it is.

Where the contribution figure comes from

The whole calculation rests on one number: what a subscriber contributes each month after the costs that come with them. Price is the start, not the answer. Where the subscription is sold changes it more than most teams expect.

Then subtract what else scales with each subscriber: refunds and chargebacks, the cost of serving them, such as hosting, content licensing, or support, and any sales tax your price includes. What remains is contribution. The retention table above assumes $12, which is roughly what the 15% store case leaves once a small allowance for refunds and serving costs comes off.

The same subscriber can therefore be worth several dollars a month more through one checkout than another. That difference compounds across every month they stay, and it changes what you can afford to pay for them. Work out contribution per checkout before comparing channels that send people to different ones.

Payback by channel

Channels do not only cost different amounts. They bring customers who stay for different lengths of time. Here are two illustrative channels at the same $12 monthly contribution. Channel A is cheaper and its cohorts decay faster, falling to 25% by month twelve. Channel B costs more and holds, ending month twelve at 44%.

Both channels pay back in the same month. By the end of the first year, the more expensive one has returned about 43% more over its cost, and the gap keeps widening for as long as the curves stay apart. A report that ranked these channels on cost per subscriber would move budget in exactly the wrong direction.

Reading cohorts before they have aged

The difficulty is that budget decisions cannot wait twelve months. The practical answer is to find the earliest point in your own history that predicts the later curve. For many subscription products that is the first renewal: cohorts that keep more people into month two usually keep more people after it. Check that relationship against your older cohorts before relying on it, because it is a pattern in your data, not a law.

• →Compare channels on month-two retention as the first read, and on month three or four as confirmation.
• →Mark projected values as projected in every report, and replace them with measured ones as cohorts age.
• →Keep cohorts small enough to be comparable, usually one month of starts, and large enough to be stable. A cohort of forty subscribers will swing on a handful of cancellations.
• →Recheck the prediction every quarter. A product change or a new offer can break the relationship between early and late retention.

From payback to a maximum cost per subscriber

1. Decide the payback period the business can fund. This is a cash decision for the owner, not a marketing one.
2. Read the cumulative contribution at that month from your retention table. In the example, a six-month payback allows about $45.72 per subscriber.
3. Work backward to cost per trial by multiplying by the trial-to-paid rate. At 30%, a $45.72 subscriber means about $13.72 per trial.
4. Set platform targets from those numbers, then recheck them by channel as cohort data arrives, because channels retain differently.

Common mistakes

• →Using average revenue per user across the whole base instead of the contribution of a newly acquired cohort.
• →Ignoring store fees, payment costs, refunds, and chargebacks, which can take a meaningful share of the price.
• →Averaging retention across channels, so a channel with weak cohorts borrows the reputation of a strong one.
• →Treating a projected curve as a measured one. Early cohorts are estimates until they have aged. Label them.
• →Judging payback on trial cost, which leaves out the most important division in the calculation.

The CAC payback calculator runs the same arithmetic for a constant churn rate. The table above is better once you have a real curve, because early churn is rarely constant. For the rest of the acquisition picture, start with the guide to paid acquisition for subscriptions.