The biggest mistake subscription businesses make when running paid acquisition: evaluating campaigns the same way a direct sales business would.
A customer paying €9.99 per month generates €9.99 in revenue on day one. If you spent €40 to acquire them, your day-one ROAS is 0.25x. By direct sales logic, that's a disaster. By subscription logic, it might be a perfectly sound investment — depending on how long they stay.
The metrics are the same. What they mean is completely different.
For a one-time purchase product, the entire customer value is realized at the moment of purchase. Your CPA either makes sense against your margin or it doesn't. The math is immediate.
For a subscription product, value accrues over time. A customer who pays €9.99 per month for 12 months generates €119.88. A customer who churns after month two generates €19.98. Same acquisition cost. Radically different return.
This means subscription businesses can't evaluate acquisition cost against first-month revenue. They need to evaluate it against predicted lifetime value — which requires knowing their retention.
Before setting any CPA target for a subscription product, calculate your monthly churn rate.
Monthly churn rate = Cancelled subscribers / Total subscribers × 100
If your monthly churn is 5%, your average customer stays for 20 months. At €9.99 per month, that's €199.80 in lifetime revenue per customer.
If your monthly churn is 15%, your average customer stays for 6.7 months. Same price point — €66.93 in lifetime revenue.
Same acquisition cost, same monthly price, completely different sustainable CPA. At 5% churn you can afford to pay significantly more to acquire a customer than at 15% churn. Running paid acquisition without knowing your churn rate is like driving without knowing how much fuel you have.
Trial start rate: the percentage of landing page visitors who start a trial. Benchmark: 3-15% depending on price point and product. This is your top-of-funnel conversion metric — it tells you whether your ad-to-landing page message is aligned.
Trial to paid conversion rate: the percentage of trial users who convert to paid. This lives outside your ad platform — it's a product and onboarding metric. But it directly affects your sustainable CPA. If your trial-to-paid rate is 20%, you need 5 trial starts to get 1 paid subscriber. If your ad platform CPA for trial starts is €8, your true CPA for paid subscribers is €40.
D7 and D30 retention: the percentage of users still active 7 and 30 days after first use. These early retention signals predict long-term retention and therefore LTV. A cohort with low D7 retention will churn heavily regardless of what you do after.
Cohort ROAS: revenue generated from a cohort of acquired users over time. D7 ROAS, D30 ROAS, D90 ROAS. Most subscription businesses set a payback period target — 6 months is common — and evaluate campaigns against whether they're on track to pay back within that window.
MRR from paid acquisition: monthly recurring revenue added from campaigns. Track this alongside CAC to understand whether paid acquisition is growing the business or just offsetting churn.
CPA: cost per acquisition. Simple and direct. Must be well below your gross margin to be sustainable. At a €29 product with 90% digital margin (€26.10 gross profit), a sustainable CPA is roughly €13-18.
ROAS: revenue divided by spend. For digital products with high margins, a ROAS of 2.5x is the minimum for a profitable campaign. 4x+ is strong. Unlike subscription products, this metric is calculable immediately — no need to wait for cohort data.
Gross margin per sale: revenue minus payment processing fees, platform fees, and any delivery costs. For a digital download this is typically 85-92% of revenue. Your CPA target should be no more than 50-70% of your gross margin per sale to maintain healthy unit economics.
Payback period: for one-time purchase products, this is immediate. Either the margin covers the acquisition cost or it doesn't. There's no compounding LTV to offset a negative first transaction.
LTV for digital products: not zero, but harder to calculate. Consider repeat purchases if you have multiple products, upsell potential to higher-ticket services, and referral value. Even without repeat purchases, a buyer who refers one friend effectively doubles their LTV.
For subscription products:
Maximum CPA = (Monthly revenue per user / Monthly churn rate) × Target payback ratio
Example: €19.99 monthly price, 8% monthly churn, 40% target payback ratio in month one. LTV = €19.99 / 0.08 = €249.88. Maximum CPA = €249.88 × 0.40 = €99.95.
That means you can afford to pay up to €99.95 to acquire a subscriber and still hit a 6-month payback. Most teams set their CPA targets far too low because they're looking at month-one revenue instead of LTV.
For one-time purchase products:
Maximum CPA = Product price × Gross margin % × 0.6
Example: €49 product, 90% margin, conservative 60% ratio. Maximum CPA = €49 × 0.90 × 0.60 = €26.46.
Run campaigns above this CPA and you're subsidising customer acquisition from other revenue. Below it, you're profitable on paid acquisition from day one.
This difference in value realization also affects how you should optimize campaigns on Meta and TikTok.
For one-time purchase products: optimize for Purchase. The event happens, the value is realized, the signal is clean.
For subscription products: optimizing for subscription purchase often produces too few events for the algorithm to learn from. Instead, optimize for the highest-frequency event that's strongly correlated with paid conversion — usually trial start or a specific onboarding milestone.
Then monitor cohort quality in your analytics separately. The platform will optimize for volume of your chosen event. You need to ensure that volume is coming from users who actually stick around.
For subscription: weekly new subscribers from paid, trial-to-paid conversion rate, D30 retention by acquisition cohort, MRR added from paid, blended CAC.
For one-time purchase: daily spend, CPA by campaign and creative, ROAS, initiated checkout to purchase rate, total revenue from paid.
The instinct to use the same dashboard for both is understandable — the metrics look similar. The interpretation is completely different.
Need help with paid acquisition for your app?
Book a call →Meta advertising looks meaningfully different in 2026 than it did two years ago. Here's what's actually working for digital product campaigns — from Andromeda to Advantage+ to attribution changes.
Most teams that try TikTok ads fail for the same reason — they run it like Meta. Here's what's actually different about the platform and what changes when you get it right.