To set a target CAC for a subscription app, work backwards from what a paying subscriber is worth, not forwards from what other apps pay. Take net revenue per subscriber over the longest period you have actually observed, subtract variable costs, decide what share of that contribution you are willing to spend on growth, then check the answer against the cash your business can tie up at the spend you plan. Your target CAC is the lowest of those ceilings, applied to the next block of spend rather than the account average. There is no category benchmark that can do this for you, because two apps with identical cohorts can rightly choose different targets.
This guide sets out the back-solve step by step, with a worked example built on the same hypothetical app we used for CAC payback period, so the numbers carry across. Every figure in the worked example is hypothetical and labelled as such. Published figures come from RevenueCat's State of Subscription Apps 2026 and from Apple's and Google's own fee documentation.
What a target CAC is, and what it is not
Subscriber acquisition cost, or CAC, is the acquisition spend required to win one paying subscriber. Our guide to calculating subscription app CAC covers how to measure it: paid versus blended, media only versus fully loaded, and why it must be cohorted by install date. A target CAC is a different thing. Measured CAC describes the past. A target CAC is a decision about the future: the most you are prepared to pay for the next paying subscriber.
It is also not the same as break-even. Our guide to break-even CPI for subscription apps shows where an install stops losing money by a chosen date. That is one input. A target adds two judgements break-even leaves out: how much of the subscriber's value you want to keep as margin, and whether your bank balance can survive the wait. Most of the mistakes we see in target setting come from treating one of those three as if it were the whole answer.
Why LTV divided by three is the wrong starting point
The method most current guides give is to estimate lifetime value, divide by a target LTV:CAC ratio of three, and call the result your maximum CAC. The ratio comes from B2B SaaS, and it has three problems when a subscription app uses it.
- The lifetime is a forecast. Most apps have not observed a cohort for as long as the LTV they are dividing. A ratio applied to a guess is still a guess.
- The value is often gross. Subscription dashboards commonly report revenue before store commission and tax. For a UK App Store subscriber in their first year, the developer receives about 58% of what the customer paid.
- It has no clock. A 3:1 ratio earned over three years and one earned over one year produce the same target, and very different cash positions.
Here is what that looks like on one cohort. Our hypothetical app's 24 month revenue per subscriber, read gross off a dashboard, is £49.15. Divide by three and the "conservative" target is £16.38. After six observed payments, the same subscribers have contributed £15.37 net of store fees, VAT, refunds and variable costs. The rule that sounds cautious would set a target above anything the app has actually seen its subscribers pay back. We go further into the ratio's strengths and weaknesses in LTV:CAC for subscription apps. The short version is that it is a fair question for an investor and a poor way to set a bid.
The target CAC formula: three ceilings, take the lowest
Value ceiling = net revenue per paying subscriber at your observed horizon × (1 − variable cost share) × the share of contribution you will spend on acquisition
Cash ceiling = the highest CAC whose cash low point, at your planned monthly spend and including store payout timing, fits inside the money you can commit
Target CAC = the lower of the two, applied to marginal spend, not the account average
The worked example below walks through each input. The app is hypothetical. It sells a £9.99 monthly subscription on the UK App Store, including 20% VAT. Apple keeps 30% in a subscriber's first year and 15% after a year of paid service, per Apple's subscription terms, so each payment nets £5.83 in year one and £7.08 after. We assume 3% of first payments are refunded. Of every 100 subscribers, 57 make a second payment, 23 are still paying after six renewals and 11 after twelve, which sits inside RevenueCat's 2026 medians of 53% to 61% for the first monthly renewal and 6% to 14% for monthly retention after a year, without being anyone's real curve.
Step 1: net subscriber value at the horizon you have actually observed
Start with cumulative net revenue per paying subscriber, cohort by cohort, and stop at the oldest cohort age you have real data for. If your oldest paid cohort is seven months old, your horizon is six or seven payments, not 24, whatever your forecast says.
| Horizon (hypothetical app) | Net revenue per subscriber | Contribution after 10% variable costs |
|---|---|---|
| Six payments (months 0 to 5) | £17.07 | £15.37 |
| Twelve payments (months 0 to 11) | £22.90 | £20.61 |
| Twenty-four payments (months 0 to 23) | £29.70 | £26.73 |
Two points about that table. First, value keeps rising with the horizon, so an app's target CAC can legitimately rise as its cohorts mature and prove the later renewals. That is the honest way to earn a higher bid: by observing it. Second, month two onwards is worth far less per month than month one, because close to half of monthly subscribers never pay a second time. The twelve payments after month 11 add £6.80 of net revenue per original subscriber, barely more than the first payment's £5.65.
Step 2: subtract the costs that scale with subscribers
Store fees are not the only cost that rises with every new subscriber. Hosting, content licences, customer support and, increasingly, AI model inference all scale with usage. We use 10% of net revenue in the example as a hypothetical figure for a conventional app. Your own figure belongs here, not ours.
For AI apps the line can be large enough to change the answer. At a hypothetical 30% variable cost, the twelve payment contribution in our example falls from £20.61 to £16.03, cutting the value ceiling by about a fifth before any other assumption moves. RevenueCat's 2026 report adds a retention warning on top: its median twelve month retention for AI apps is 6.1% on monthly plans against 9.5% for non-AI apps, and 21.1% against 30.7% on annual plans. The same report finds AI apps earn 41% more revenue per payer, so the higher price can cover some of this, but only if both lines are in the model.
Step 3: decide how much of that value you are willing to spend
This is the growth appetite, and it is a business decision, not a calculation. Spending 100% of contribution at your horizon means acquisition breaks even by that date and every pound of profit comes from renewals after it. Spending 80% means you keep a fifth of the observed contribution as margin by the horizon. Spending more than 100% means you are betting on renewals you have not yet seen.
There is one legitimate reason to pay above the value ceiling for a while: learning. New concepts, new audiences and new markets cost more before the account finds the people who convert. That spend is a research budget. Size it separately and cap it, as set out in our guide to the Meta ads testing budget, rather than quietly raising the target CAC to absorb it.
Step 4: check the target against your cash
A CAC that pays back on paper can still run out of money in practice. You pay for media as it runs, but the stores do not pay you when the subscriber pays them. Apple pays proceeds within 45 days of the end of the fiscal month in which the transaction completed, and Google Play pays around the 15th of the following month. We model the gap as two months, as in the payback guide.
Here is the cash low point for the hypothetical app spending a steady £30,000 a month on Meta, at different CACs, with the 10% variable cost included.
| CAC per subscriber | Payback month (net, after variable costs) | Cash low point at £30,000 a month |
|---|---|---|
| £12 | Month 3 | About £91,000 |
| £15 | Month 5 | About £110,000 |
| £18 | Month 8 | About £134,000 |
| £20 | Month 11 | About £154,000 |
Read the table in reverse to get the cash ceiling. With £100,000 available to finance growth, the highest CAC this app can run at £30,000 a month is about £13.45. With £150,000 it is about £19.60, and with £250,000 about £26.53. Note what does not move: even at a very low CAC, two months of media are paid for before the first payout lands, so this app needs at least £60,000 of working capital at £30,000 a month whatever its CAC. If your planned spend needs more cash than you have, the answer is a lower target or a lower spend, not a more optimistic retention curve.
Worked example: same cohorts, three different target CACs
Now apply the method to three hypothetical apps with exactly the same subscribers and the same £30,000 monthly budget. Only their finances differ.
| App (hypothetical) | Value ceiling | Cash ceiling | Target CAC |
|---|---|---|---|
| A: self-funded, £100,000 available to finance growth, six months of cohorts | £15.37 × 80% = £12.30 | £13.45 | £12 |
| B: funded, 18 months of runway, £250,000 earmarked, twelve months of cohorts | £20.61 × 100% = £20.61 | £26.53 | £20 |
| C: raising in six to nine months, £150,000 available, twelve months of cohorts | £15.37 × 100% = £15.37 (six payment horizon, so cohorts visibly pay back before the round) | £19.60 | £15 |
App B can pay two thirds more per subscriber than App A for the same customer, and both are right. App A is bound by margin: it wants to keep a fifth of what subscribers contribute within six months. App B can afford to break even at twelve payments because it has observed twelve and has the cash to wait. App C has twelve months of data but chooses a six payment horizon anyway, because the cohorts it buys now need to have visibly paid back before investors look at them.
This is why a published "good CAC" for your category is close to useless as a target. It cannot know your horizon, your margin requirement or your bank balance, which are the three things that separate £12 from £20 here.
Step 5: apply the target to marginal spend, not the average
A target CAC is a limit on what the next subscriber costs. The account average will hide a breach. Suppose App C spends £20,000 a month at an average CAC of £12, then adds £10,000 that buys subscribers at £20. The blended CAC across £30,000 is £13.85, comfortably inside its £15 target. The last £10,000 is a third over it.
So raise budgets in steps and judge each step on the cohort it bought. If the marginal CAC is over target, the extra spend is buying subscribers you have decided are not worth it, however healthy the dashboard looks. This is the single most common way we see a sensible target CAC become a meaningless one.
Turn the target into numbers a Meta account can hit
Nobody can steer a campaign on a cost per subscriber that takes weeks to read, especially on iOS where attributed conversions arrive late. Translate the target down the funnel. Target cost per trial is target CAC multiplied by your trial-to-paid rate. Target CPI is target CAC multiplied by both install-to-trial and trial-to-paid. Using hypothetical rates of 12% install-to-trial and 35% trial-to-paid:
| Target CAC | Target cost per trial (× 35%) | Target CPI (× 12% × 35%) |
|---|---|---|
| £12 | £4.20 | £0.50 |
| £15 | £5.25 | £0.63 |
| £20 | £7.00 | £0.84 |
Two consequences follow. First, a £9.99 monthly app converting 4.2% of installs into payers needs installs well under a pound, which is the economic reality many UK monthly-only apps run into on paid social. The levers are the conversion rates and the price, not the bid. Our install-to-trial and trial-to-paid benchmarks show where your rates sit against category medians. Second, the derived targets only hold if the conversion rates come from your paid cohorts. Organic users usually convert differently, and a CPI target built on blended rates will be wrong in a direction you will not like. Our comparison of CPI, CPA and CAC for subscription apps covers which of these numbers to act on, and when each becomes readable.
Set targets by store and market, not one per app
Fees differ by store, so the value ceiling does too. Google Play charges 15% on auto-renewing subscriptions, made up of a 10% service fee and a 5% billing fee where Google Play Billing is used, according to Google's service fee documentation. Run the same retention curve through Google Play in the UK and the twelve payment contribution rises from £20.61 to £25.03, about 21% more. Do not assume the curve really is the same, though. RevenueCat's 2026 report finds involuntary billing failures make up 31% of subscription cancellations on Google Play against 14% on the App Store, so Android retention often needs its own line. We cover the cost side in iOS versus Android CPI.
Markets change the inputs as well. VAT rates differ by country, and from 1 October 2026 Apple's commission on EU storefronts is 26% for In-App Purchase, falling to 15% for auto-renewing subscriptions after their first year. Each market you buy in deserves its own value ceiling.
What about annual plans?
Annual plans make the value ceiling simpler and more brittle. In our example a £59.99 annual subscription nets £33.94 on its first payment after Apple's 30%, VAT and refunds, or £30.55 after the 10% variable cost. If you will not count renewals you have not observed, that first payment is your entire value ceiling, and it is collected on day one. The next pound of value does not arrive for twelve months. We will compare plan durations properly later in this series. For now, the point is that an app with a mix of plans needs a target built on its own plan mix, because the same target CAC can be generous for annual buyers and ruinous for monthly ones.
What we look at when a client gives us a target CAC
When a subscription app hands us a target, we do not argue with the number first. We ask where it came from. These are the questions, in order.
- What horizon is it built on, and how old is your oldest paid cohort?
- Is the value net of store commission, sales tax, refunds and variable costs?
- What share of contribution have you decided to spend, and who decided it?
- What is the cash low point at the spend you want, including store payout timing?
- Is it a marginal target, or an average one that the next budget increase will quietly break?
The answers usually matter more than the number. A target that survives all five is one we can build an account plan around. One that does not is often too high, which is the expensive failure, but sometimes too low, which starves a business that could afford to grow faster.
The other thing we look at is which side of the target is moveable. Media efficiency moves CAC, but so does what happens after the install. In the Steps & Beasts case study, heavy creative testing alongside onboarding redesigns lifted revenue 1112% and active subscriptions 118%. The case study does not publish a CAC or a target, and we will not invent one. What it shows is that more of the people the ads bring in can be turned into subscribers, which lowers the cost per subscriber without the media getting any cheaper.
Frequently asked questions
What is a target CAC for a subscription app?
A target CAC is the most you are prepared to pay, in acquisition cost, for one new paying subscriber on your next block of ad spend. It is a decision rather than a measurement. It should be set from net subscriber value over the period you have actually observed, the share of that value you are willing to spend on growth, and the cash low point your business can fund at the spend you plan.
How do you calculate the maximum CAC for a subscription app?
Work backwards in five steps. Find net revenue per paying subscriber at the oldest cohort age you have real data for, after store commission, sales tax and refunds. Subtract variable costs such as hosting, support and AI inference to get contribution. Multiply by the share of contribution you are willing to spend on acquisition. Check the result against the cash low point at your planned spend, including store payout timing. The lowest of those ceilings is your maximum CAC.
Should target CAC be LTV divided by three?
Not for most subscription apps. The 3:1 rule comes from B2B SaaS and divides a forecast lifetime value that is often gross of store fees and longer than any cohort the app has observed. In our hypothetical worked example, a 24 month gross LTV divided by three gave £16.38, which looks conservative but is higher than the £15.37 the app's subscribers had actually contributed after six observed payments.
Should my target CAC be different on iOS and Android?
Usually, yes. Google Play takes 15% on auto-renewing subscriptions (a 10% service fee plus a 5% billing fee where Google Play Billing is used), while Apple takes 30% in a subscriber's first year. With identical retention, a UK Android subscriber in our example was worth about 21% more over twelve payments. Retention is rarely identical, though: RevenueCat's 2026 report finds involuntary billing failures account for 31% of subscription cancellations on Google Play against 14% on the App Store, so use each store's own cohort curve.
Is target CAC the same as target cost per trial or target CPI?
No, but they are linked. Target cost per trial is target CAC multiplied by your trial-to-paid rate, and target CPI is target CAC multiplied by your install-to-trial and trial-to-paid rates. In our example a £15 target CAC with a 35% trial-to-paid rate allows £5.25 per trial, and with a 12% install-to-trial rate it allows about £0.63 per install. Those derived numbers are what a Meta account can be judged on within a week.
How often should a subscription app reset its target CAC?
Whenever one of its inputs changes: a new month of cohort data extends the horizon you have observed, a price or plan-mix change moves net value, a new market or store changes fees and retention, or your cash position changes. Store fees change too. Apple's commission on EU storefronts moved to 26%, and 15% for auto-renewing subscriptions after their first year, from 1 October 2026. A target CAC that has not been revisited for a quarter is usually a historical figure, not a target.
The short version
A target CAC is chosen, not looked up. Build it from net subscriber value at the horizon you have actually observed, minus the costs that scale with subscribers, multiplied by the share you are willing to spend on growth. Then check it against the cash low point at your planned spend. Take the lowest ceiling, apply it to marginal spend, and translate it into a cost per trial and a CPI the account can be judged on in a week. In our hypothetical example the same subscribers justified targets from £12 to £20 depending only on margin, horizon and cash, and the popular LTV-divided-by-three rule landed above anything the cohorts had paid back.
For how CAC fits into the wider growth system, see our subscription app marketing strategy guide. If you want Meta acquisition run against a target built this way, our consumer apps page explains how we work with subscription apps, and if your app is already spending, our pricing is published.