The Mobile Monetization Decision Tree: Ads vs IAP vs Subscription vs One‑Time — What to Run First (with Experiments)

Before you debate ad networks or paywalls, run four rapid signal checks that determine which models deserve a real experiment. These checks are product-led and measurable: frequency (how often users open the app), session depth (how long and what they do), willingness-to-pay signals (settings toggles, advanced features used), and audience type (B2B/professional vs casual/utility).

If frequency is low (monthly opens) and session depth shallow, subscriptions rarely work; ads or one-time purchases are better fits. If users open daily with high engagement and there are clear premium workflows, subscriptions or IAPs can scale. Use logs and a 7‑day cohort to compute these signals — you’ll need them to power the experiments below.

Use this short heuristic tree: if the app is utility/productivity with recurring value and daily/weekly habit, start with a subscription trial. If the app is casual or low-frequency but has micro‑utility moments (stickers, filters, boosters), start with IAPs. If usage is broad, free, and you can insert rewarded or native placements without destroying UX, test ads first. If the app is a one-off tool (single useful export or file conversion), test a one-time paywall.

These are heuristics, not commandments — the next step is experiments to validate the chosen path. For each branch you should have a minimal revenue hypothesis (e.g., we expect a 1% subscription conversion at $4.99/mo on iOS) and an experiment plan that can falsify it fast.

Run these experiments in parallel where possible. Each is designed to be low-implementation: server-side flags, feature gates, remote config and soft paywalls rather than full billing flows. You’ll be testing signal, not polishing checkout UX.

Keep sample sizes pragmatic: for conversion signals you’ll often need a few thousand exposures to detect percent-level conversion; if you don’t have that, rely on intent proxies (clicks to a pay prompt, trial starts) and iterate.

Use these conservative priors to decide whether to double down: for basic ad monetization expect ARPDAU (ad-only) around $0.02–$0.15 depending on region and format; rewarded video and interstitials lift eCPM substantially. For subscriptions, expect free-trial-to-paid conversion in the 1–5% range for consumer apps (higher for productivity & niche verticals). For IAPs, initial conversion usually sits between 0.5–3% with ARPPU depending on price points.

Translate these into simple stop/go numbers: if your projected ARPU (based on experiment CTR/CVR and eCPM/price) does not exceed your cost per acquisition (CPA) or the runway-adjusted target LTV within 6 weeks, stop and test another model. Always project LTV = ARPU / churn rate for subscriptions; use cohort retention to estimate 3‑month LTV.

Use three KPIs: acquisition-to-revenue conversion (CTR→CVR), ARPU (platform + region segmented), and retention impact (D1, D7, D30). Run survival‑aware experiments (or use Kaplan‑Meier/Cox models if you can) to properly measure retention and monetization trade-offs rather than just short-term spikes.

Once a model passes stop/go thresholds, instrument a production-priced checkout and scale while improving unit economics: implement mediation and ad bidding for ads, subscription introductory offers and pricing tiers for subs, or better bundle architecture for IAPs. Always keep a holdout group to detect long-term retention degradation after scaling.