An install is only one stage of an app campaign. Someone may download the app without opening it, completing onboarding, registering or making a purchase. Treating all of these as one generic “conversion” metric hides where the funnel fails and what the advertiser actually pays for. Mobile app campaigns on Meta Ads and Google Ads require separate event configurations and attribution reconciliation.
Define the event dictionary
Install is an installation under the agreed attribution rule; first open is the first app launch; signup is registration; onboarding completed is a completed first-use flow; purchase or subscription is a payment event. Developers and trackers may name events differently. Agree on definitions and uniqueness rules.
Map the funnel
Impressions → clicks → store visit → install → first open → signup → payable action. Track a conversion rate between each transition. If 2,000 installs produce 1,600 first opens and 400 registrations, install-to-open is 80%, open-to-signup is 25%, and install-to-signup is 20%.
Registrations are not revenue when the program pays for first purchase.
An install is only one stage of an app campaign. Someone may download the app without opening it, completing onboarding, registering or making a purchase. Treating all of these as one generic “conversion” metric hides where the funnel fails and what the advertiser actually pays for. Mobile app campaigns on Meta Ads and Google Ads require separate event configurations and attribution reconciliation.
Define the event dictionary
Install is an installation under the agreed attribution rule; first open is the first app launch; signup is registration; onboarding completed is a completed first-use flow; purchase or subscription is a payment event. Developers and trackers may name events differently. Agree on definitions and uniqueness rules.
Map the funnel
Impressions → clicks → store visit → install → first open → signup → payable action. Track a conversion rate between each transition. If 2,000 installs produce 1,600 first opens and 400 registrations, install-to-open is 80%, open-to-signup is 25%, and install-to-signup is 20%.
Registrations are not revenue when the program pays for first purchase.
The ad platform, mobile measurement partner and affiliate network may report different numbers due to attribution windows, deduplication, postback delays and privacy requirements. Apple explains its privacy-preserving AdAttributionKit, which delivers postbacks with potentially limited detail.
Compare event definitions, time zones, install dates and action dates—not just total counts.
Pre-launch checklist
1. Supported OS and GEO. 2. Correct store URL. 3. Configured MMP or agreed measurement method. 4. Payable event. 5. Test postback. 6. Attribution window. 7. Reinstall and duplicate rules. 8. Data protection.
Economics example
A campaign spends $800 for 2,000 validated installs, giving $0.40 CPI. One hundred sixty users complete the target event, but only 140 are approved at $7. Revenue is $980 and the result before other costs is +$180. Counting all 160 events as payable would exaggerate income.
App attribution across Meta Ads and Google App campaigns
Meta Ads can measure installs and in-app events through an SDK or an agreed mobile measurement partner (MMP); verify app linkage, permitted events and Facebook/Instagram attribution. Google App campaigns use their own measurement and imported event signals. Developers, the MMP and the affiliate network must agree on the definitions of install, first open, signup and purchase or CPI/CPE comparisons become misleading.
On iOS, allow for privacy-preserving attribution and Apple's AdAttributionKit, including delayed and limited postbacks. A single install does not become two paid events because both Meta and Google claim it. Report by OS, country, app version, channel and mature cohort; calculate earnings from advertiser-approved events.
Reporting
Segment by campaign, OS, country, app version and install cohort. Give cohorts time to complete the required event. If the funnel changes after an app release, verify event implementation before changing media budgets.
Understand attribution
The ad platform, mobile measurement partner and affiliate network may report different numbers due to attribution windows, deduplication, postback delays and privacy requirements. Apple explains its privacy-preserving AdAttributionKit, which delivers postbacks with potentially limited detail.
Compare event definitions, time zones, install dates and action dates—not just total counts.
Pre-launch checklist
1. Supported OS and GEO. 2. Correct store URL. 3. Configured MMP or agreed measurement method. 4. Payable event. 5. Test postback. 6. Attribution window. 7. Reinstall and duplicate rules. 8. Data protection.
Economics example
A campaign spends $800 for 2,000 validated installs, giving $0.40 CPI. One hundred sixty users complete the target event, but only 140 are approved at $7. Revenue is $980 and the result before other costs is +$180. Counting all 160 events as payable would exaggerate income.
App attribution across Meta Ads and Google App campaigns
Meta Ads can measure installs and in-app events through an SDK or an agreed mobile measurement partner (MMP); verify app linkage, permitted events and Facebook/Instagram attribution. Google App campaigns use their own measurement and imported event signals. Developers, the MMP and the affiliate network must agree on the definitions of install, first open, signup and purchase or CPI/CPE comparisons become misleading.
On iOS, allow for privacy-preserving attribution and Apple's AdAttributionKit, including delayed and limited postbacks. A single install does not become two paid events because both Meta and Google claim it. Report by OS, country, app version, channel and mature cohort; calculate earnings from advertiser-approved events.
Reporting
Segment by campaign, OS, country, app version and install cohort. Give cohorts time to complete the required event. If the funnel changes after an app release, verify event implementation before changing media budgets.