Branch Mobile Attribution: Keep, Add, or Replace?
Branch mobile attribution is enough when channels, events, windows, and links fit; otherwise test an added MMP or replacement, including Airbridge Core.
Keep, add, or replace Branch?
- Keep Branch when its configured channels, events, attribution windows, and link journeys meet the decisions your team needs to make.
- Add another mobile measurement partner (MMP) only for a documented requirement Branch does not satisfy, then assign one source of truth to each report.
- Replace Branch when one platform passes the same attribution and routing tests your live links depend on.
- Airbridge Core is an integrated option when its four paid channels, link features, web-to-app attribution, and reporting fit your needs.
Branch mobile attribution is enough when the Branch setup you run today can measure your required paid channels, app events, and journeys within the attribution rules you intend to use. Keep Branch in that case; add a second measurement partner only for a specific unmet requirement, or replace the stack when one alternative passes both the measurement and link-routing checks.
| Path | Choose it when | What the team does next |
|---|---|---|
| Keep Branch | Configured Branch attribution covers the required channels, installs, opens, in-app events, and link paths. | Fix event mapping, windows, or link settings that fail a defined test. |
| Add a second MMP | One named requirement, such as a reporting output or event workflow, remains unmet after you verify the Branch configuration. | Give the second tool that job, assign report ownership, and avoid counting the same result twice. |
| Replace Branch | A single alternative measures required campaigns and events, routes every important link, and fits the team's cost and data needs. | Test it alongside the current routes, then move link paths in controlled groups. |
What counts as a real attribution gap?
A real gap is a decision your current setup cannot support after you have checked event delivery, attribution windows, consent fields, and link behavior. For example, “we cannot tie trial starts to paid channels in the report we use” is testable.
Check channels against the campaigns you run
Write down the ad channels whose attributed installs and downstream outcomes you need to compare. Then verify that each channel is configured to send the touchpoints or events required for the report. A dashboard with several channel names still needs the matching integration, campaign setup, and app-side events to give the team usable results.
Check events and revenue from the app
Subscription teams usually need to separate an install from the events that show progress toward revenue. List the event names used for trial start, subscription start, renewal, cancellation, and any product action that predicts conversion. Confirm the app or server sends each event once with the right user or transaction context, then check that the report displays it in the way the team expects.
For the events the team actually uses, inspect the payload and verify that the transaction value and currency reach the expected report. Branch’s Attribution settings and Events API describe different parts of the setup; keep report eligibility separate from successful event delivery.
Check attribution windows before judging a result
A window is the period after a click or impression in which a later action can receive credit. Two tools can report different attributed installs or conversions if they use different windows, even when both receive the same events. Record the window values you need for the decision, then compare those values with the live project settings.
Export the values configured for click-to-install, click-to-conversion, impression-based credit, and re-engagement from the current account. Label them with the app and configuration date. Default examples in Branch’s guide are a reference, not proof of the settings applied to a live app.
If a purchase occurs ten days after a qualifying click, a seven-day conversion window excludes that click from credit for the purchase. A longer configured window can change eligibility; test the actual event and competing touchpoints before interpreting the result as a reporting defect.
Check consent and event eligibility
For affected traffic, inspect the event data for applicable regulation and consent fields, then compare the same event, campaign, period, and consent population.
When does a second MMP add value?
A second mobile measurement partner adds value when it has a distinct job tied to a requirement Branch does not meet. For paid channels, compare each candidate's current supported-network list and included plan against the campaigns you buy.
Assign one source of truth for install attribution, subscription revenue, and each decision the team makes; give the added tool a separate job, such as a required server-side event or a particular output. Compare both tools over the same date range and event definition, and reconcile subscription purchases against the billing record.
When does replacing Branch with one platform fit?
Replacement suits a team that wants fewer moving parts and can confirm that one product covers both the required attribution work and the link journeys already in use. The candidate has to pass the same list as Branch: paid channel coverage, install and conversion events, revenue reporting, data access, and routes for links in ads, email, web pages, and onboarding. For a subscription app, test the full path from campaign click through install and trial or purchase, rather than judging on an install count alone.
Use this short fit test for an integrated option:
If an essential channel or data output falls outside the documented plan scope, a different plan or a second tool may be the better architecture. If those requirements fit, test the candidate with real campaign links and subscription events while Branch remains available for the existing routes.
How to validate the setup before changing live links
Keep the existing live routes in place until the new path returns the expected destination and measurement result across your important user journeys.
- Inventory current links and destinations. Export or list the Branch URLs used in live ads, email, web pages, QR codes, and referrals. For each URL, write the expected destination for someone with the app installed, someone installing the app, and someone opening the link on the web. Note whether the URL carries campaign or other analytics parameters.
- Write the measurement acceptance criteria. Name the required paid channels, events, currencies, and reporting view. Record click and impression windows, the re-engagement rule, and the period you will use to compare results. Include consent-field handling for users where your setup requires it.
- Choose ownership for each data path. Decide which product records install attribution, which receives subscription events, and where the team reads revenue. When testing a second MMP, give it a distinct job and set a report as authoritative for each decision. This keeps the team from adding counts from two tools as though they represented unique subscribers.
- Build a test version and exercise the routes. Test direct links for users with the app installed, deferred links after a new install, and web fallback pages. Open test links from the same kinds of placements your users encounter, such as a browser and a messaging app. On each platform and app version you support, confirm that the link reaches the expected screen and preserves the campaign parameters the report needs.
- Verify events and reports with controlled test records. Trigger each required event using a test account, then confirm the event name, value, currency, campaign, and timestamp appear in the selected reporting view. Check subscription revenue against the billing record. Compare results for the same test user and event in both tools if the test stack temporarily sends to two providers.
- Change domains and live routes in a controlled sequence. Branch's Advanced Link Configuration guide requires LIVE and TEST apps to use different customized Link Domains. If both Branch environments need a changed domain, Branch recommends changing LIVE first and using a distinct test domain. For any replacement, test the new domain and destinations with a limited group before moving links in larger sets, and keep an inventory of which URLs have changed.
- Decide from the pass/fail record. Keep Branch if it passes the required channel, event, reporting, and routing checks. Add a second MMP if a verified requirement remains unmet and the new tool passes its own test with a clear owner. Replace Branch only when the alternative passes both the measurement checks and the link scenarios that matter to current users.
The result is a decision based on observed behavior, not a second dashboard added in the hope that two sets of attributed installs will agree. Keep the report definitions, link inventory, and test results together so the next person who changes the SDK or campaign route can repeat the same checks.
Keep a link-dependency register before a replacement
Record the owner of each custom domain, the app entitlement or Android association using it, active ad and email placements, destination rules, and long-lived QR codes. Ask who can keep old links resolving after cancellation and what the service agreement permits. Approve replacement only after one owner can account for both the old published URLs and the new measurement path.
FAQS
FAQ
What is mobile attribution?
Mobile attribution connects app installs or later app events with eligible marketing touchpoints, using configured rules such as click or impression windows. It helps a team compare which campaigns receive credit for the outcomes it measures.
What does Branch.io do?
Branch's Attribution Methods documentation describes attribution of app installs, opens, and in-app activity to the marketing that drove them. Its configuration also controls attribution windows, event eligibility, and link behavior.
Should I add attribution to Branch?
Add a second MMP when a documented requirement remains unmet after you verify Branch's channels, event delivery, consent fields, and windows. Give the new tool a defined job and report owner before sending it live data.
Does a second MMP make attribution more accurate?
No. Add a second MMP when a controlled test shows that it meets a distinct requirement, and compare its results with the team's defined source of truth.
Get Started Free
See how these criteria hold up on the real thing.
