iOS subscription LTV: why the simple churn formula misleads and how to build a cohort model

The standard LTV = ARPU ÷ churn formula assumes constant churn—but iOS subscriber churn is heavily front-loaded. Here's how to build a cohort-based LTV model from App Store Connect data and use it to set better CAC targets and pricing decisions.

Most iOS subscription developers reach for the same formula when they need a lifetime value number: divide average monthly revenue per subscriber (ARPU) by your monthly churn rate. If ARPU is $9 and monthly churn is 5%, LTV is $180. Simple, fast, and — for most apps — materially wrong.

The formula isn't wrong in principle. It is the correct result for a subscriber who churns at exactly 5% every single month indefinitely. The problem is that real iOS subscribers don't behave that way. Churn is heavily front-loaded: published data from RevenueCat's annual State of Subscription Apps reports has consistently shown that a disproportionate share of monthly subscription cancellations happens in the first few billing periods. Subscribers who renew past month four or five are a different population — more engaged, less price-sensitive — and they churn at rates materially lower than the blended figure used in the denominator.

Using a single blended churn rate therefore tends to overstate the value of early-cohort subscribers (where real churn is high) while underrepresenting the stickiness of longer-tenured subscribers. The result is an LTV figure that research from Adapty and RevenueCat suggests is often 20–40% higher than what cohort data actually supports. That error compounds into inflated CAC budgets, distorted annual-vs-monthly pricing math, and acquisition channels that look more efficient than they really are.

This post explains why the formula fails, walks through a cohort-based alternative using data already in App Store Connect, and covers how the corrected number should change specific pricing and acquisition decisions. For context on the broader analytics landscape, see our guide to iOS subscription analytics: MRR, churn rate, and LTV explained.

Why the standard LTV formula breaks for iOS subscriptions

The classic formula derives from a simple probability model: if a subscriber has a fixed probability c of churning each period, the expected number of periods they stay is 1/c. Multiply by ARPU and you get LTV = ARPU ÷ c.

That model has one core assumption: churn probability is constant across all periods. In statistical terms, it assumes churn follows a geometric distribution. The implication is that a subscriber who has been paying for 18 months has the same probability of cancelling next month as a brand-new subscriber does in month two. Almost no real subscription app shows this pattern.

What actually happens in iOS subscription cohorts — documented in RevenueCat's annual State of Subscription Apps benchmarks and reinforced by Phiture's mobile growth research — is a survival curve that drops steeply in months one through three, then flattens considerably. The subscribers who reach month four have survived the "trial hangover" (people who forgot to cancel), the first "do I actually use this?" reconsideration, and the first moment they questioned the price. What remains is a core of engaged users who tend to churn at rates that can be less than half of early-period churn.

When you compute churn as a single blended number across all periods and cohort sizes, you mix the high-churn early months with the low-churn later months. The denominator in LTV = ARPU ÷ churn ends up too high, making LTV look better than cohort data supports.

Cohort-based LTV: the right mental model

The fix is to stop treating churn as a single rate and start treating subscriber value as a survival-weighted cash flow series. The math is straightforward:

• Let S(t) be the fraction of the original cohort still active at period t (the survival rate).
• Revenue attributed to the cohort in period t equals S(t) × ARPU.
• Cohort LTV = Σ [S(t) × ARPU] across all periods.

You can add a discount factor for the time value of money, but for most indie apps — where subscription lengths span one to four years and cost of capital is modest — an undiscounted 24-month sum lands within a few percent of the theoretically precise number.

The survival rate S(t) is exactly what App Store Connect's retention report gives you. If month-3 retention for a cohort is 46%, then S(3) = 0.46. Multiply by net ARPU (after Apple's commission), sum the column across periods — that is your cohort LTV estimate. No complex tooling required beyond a spreadsheet.

The cohort LTV is lower than the formula suggests for apps with significant early-period churn. It can occasionally be higher for apps where the survival curve is unusually flat — some B2B tools and utility apps with strong habitual use show this. The goal isn't to find a lower number; it's to find the right one.

What App Store Connect provides — and where it falls short

Apple's native subscription analytics give you the raw material for a basic cohort model, but they have meaningful gaps. Understanding those gaps tells you when you need a third-party tool.

For most developers, App Store Connect's native data is sufficient for a directional LTV model. The primary gap is that it tracks subscriber headcount rather than revenue — so if you have multiple price points (a monthly plan and an annual plan at different tiers), you need a blended ARPU estimate applied uniformly. That introduces error if your plan mix shifts substantially over time.

Third-party SDK platforms fill these gaps. RevenueCat's dashboard shows LTV by cohort directly, segmented by plan and territory. Adapty provides similar views. If your app already uses one of these for receipt validation, cohort revenue reporting is typically available within existing plan tiers — no additional tooling cost required. For more on territory-level revenue differences and how they feed into LTV, the AppsOps territory data surfaces regional trends by storefront.

How cohort LTV changes pricing and acquisition decisions

The reason to build a cohort LTV model is not academic. It changes the numbers that underpin two of the most consequential decisions in subscription app growth: how much to spend acquiring users, and how to structure annual versus monthly pricing.

Setting CAC targets. A widely cited rule of thumb holds that customer acquisition cost (CAC) should not exceed roughly one-third of LTV on a 12-month payback horizon. If your formula LTV says $180, that implies a CAC ceiling around $60. If cohort LTV says $105, the same rule implies $35. That difference is significant when setting bids on Apple Search Ads — and apps that calibrate CAC against inflated LTV estimates often discover the economics erode as cohorts mature past the first few months. For more on Apple Search Ads economics, see Apple Search Ads for subscription apps: campaign structure, trial optimization, and measuring ROAS.

Annual versus monthly subscriber value. One of the most common LTV errors is applying a monthly churn rate directly to annual subscribers. Annual plans create a structural 12-month floor on subscription duration — even a dissatisfied subscriber stays active until the renewal date. This makes annual LTV higher per subscriber on a per-year basis, but the right comparison is cohort renewal rate at year one, not an extrapolation of monthly churn figures. RevenueCat's data has suggested that annual subscribers tend to exhibit higher renewal rates at the one-year mark than monthly subscribers show at equivalent cumulative tenure, though conversion to annual is lower — meaning blended LTV across all subscribers depends heavily on annual plan take rate. For the conversion incentive math, see subscription pricing on iOS: monthly vs yearly conversion math.

Territory-level LTV variation. Cohort LTV varies by geography in ways that frequently surprise developers. A market with lower ARPU due to PPP-adjusted pricing may show higher monthly retention, producing comparable or higher long-run LTV than a high-ARPU market with aggressive early churn. Developers who evaluate territories by ARPU alone — and not cohort retention — often under-invest in markets that generate strong lifetime revenue. This is a structural argument for using PPP-calibrated local pricing: the lower price point reduces price-sensitive early churn, which materially extends the survival curve.

When to raise prices. If cohort LTV turns out to be higher than the formula implied — which happens for apps with unusually flat retention — you may have more headroom than you realized to invest in paid growth before optimizing price. Conversely, if cohort LTV is materially lower, raising price before fixing churn is typically the wrong order of operations. A price increase tends to steepen early-period churn among price-sensitive subscribers, compressing the survival curve further and reducing LTV even if the immediate ARPU impact looks positive.

The cohort LTV model is not a complex tool — it is a spreadsheet with the right inputs. The input that matters most is the survival curve from App Store Connect's retention report. Getting that number right, and using it consistently across CAC decisions, pricing experiments, and territory analysis, is one of the most leveraged modeling improvements available to subscription app developers who are currently relying on the formula shortcut.

Sources and further reading

• RevenueCat Blog — annual State of Subscription Apps reports with cohort benchmarks, churn rates, and LTV data by category
• Apple App Store Connect Help — official documentation including subscription analytics, retention reports, and cohort views
• Phiture — mobile growth agency publishing research on retention benchmarks, ASO, and subscription economics
• Adapty Blog — LTV calculation methods, paywall analytics, and subscription benchmarks by app category
• AppFollow Blog — App Store analytics, rating trends, and subscription performance research