CPI: Cost Per Install
The average cost of acquiring one app install through paid user acquisition campaigns.
Formula
Why it matters
CPI is the primary efficiency metric for mobile user acquisition. It tells you the cost of filling the top of your app funnel — but must be read alongside trial-to-paid conversion rate and LTV. A low CPI with high churn produces negative unit economics.
How to improve CPI
Test creatives with native/UGC aesthetic versus produced video, target lookalike audiences from high-LTV users, reduce friction in the app store listing (better screenshots, A/B test the description), and optimise bidding toward in-app events rather than installs.
iOS (US market): $2–8. Android (US): $1–4. Gaming globally: $1–4. Subscription apps: $2–10. Varies heavily by category and geo.
Prooflytics tracks CPI automatically from your connected sources and flags it in your daily briefing when it moves significantly.
Book a demoCPI calculator
Calculate Cost Per Install
CPI
CPI = Ad Spend ÷ Installs
—
Frequently asked questions
What is a good CPI for iOS apps?
For iOS apps in the US, a CPI of $2–5 is typical for most categories. Subscription-based apps (fitness, finance, productivity) often see $4–10 because they target higher-intent users. Gaming apps can achieve $1–3 at scale with optimised creative. Always compare CPI to trial-to-paid rate and LTV — a $6 CPI that converts at 40% trial-to-paid is more profitable than a $2 CPI that converts at 5%.
How is CPI different from CPA?
CPI counts the cost per app install (a download event). CPA (Cost Per Acquisition) counts the cost per meaningful conversion — typically a trial start, subscription purchase, or in-app purchase. CPA is the more strategic metric because it accounts for install-to-conversion funnel quality.
What is the difference between CPI and eCPI?
eCPI (effective Cost Per Install) is a normalised metric used when comparing campaigns with different buying models (CPM, CPC, CPA). eCPI = Total Spend ÷ Total Installs, regardless of how the campaign was bought. It enables apples-to-apples comparison across ad networks.