AppsFlyer attribution, or the attribution from any mobile measurement tool, tells you which campaign produced an install or an in-app event. It cannot tell you which of those users became a paying client, because that happens in your CRM. And when both systems hold a view, they often disagree: the app tool credits a paid install campaign while the CRM credits organic search, because that is how the same person first filled in a form a month earlier. This article answers how to carry app-side source data onto contacts in the levelpartner.ai platform, store both views side by side, and decide which one drives budget.
The question behind the search
People who search for AppsFlyer analytics or attribution usually have an app and want to know which campaigns bring users who matter. Most service businesses use their app for one of three jobs: a client portal, a calculator or lead magnet, or a booking route. In each case the app is a step towards a sales conversation, not the sale. App analytics shows sessions, retention and events. The board wants won work by acquisition channel.
So there are two jobs. The first is a handover: when an app user identifies themselves by signing up, requesting a quote or booking, the source the app knows about should arrive on the CRM contact. The second is reconciliation: deciding what to do when that source contradicts the one the CRM recorded itself.
Why the two systems disagree
Each system sees only the touches inside its own walls. The app tool knows the click or view that came before the install. The platform knows the first form, call, booking or chat it recorded, and the latest one. A person might find you through search, read three articles, see a social advert, install the app and then book through it. Four touches, three systems, and every system claims the client.
The answer is not to choose a winner in advance. Store each system's claim on the contact, label it clearly, and apply one agreed rule for budget decisions.
What to check in any mobile measurement tool
This article does not describe third-party products, so treat this as a checklist for whichever tool you use:
- Can it pass campaign, media source and click identifiers to your own systems when a user converts?
- Can it send that data server to server, rather than only into its own dashboard?
- Does it identify users by something you also hold on the CRM contact, such as an email or a customer number?
- Does your privacy notice and the app's consent flow cover sharing that data with your CRM?
If the answers are yes, the rest is configuration on the platform side plus a short piece of developer work.
Before you start
- An app event that genuinely represents a lead, such as account created, quote requested or booking made.
- A developer, or the app tool's support team, able to send that event and its attribution values to your CRM.
- The platform's own first and latest attribution already recording from forms, calendars and chat.
- A pipeline where app leads are tracked through to won or lost, with values.
- A written rule, agreed by marketing and sales leadership, for which source wins in budget reports.
How to set it up
- Write down the rule for which source wins budget decisions.
- Go to Settings, then Custom Fields, and stay on the Fields tab.
- Select Create field and choose the Field type.
- Enter a clear Field name, such as App attributed campaign.
- Check the Live preview, then click Create custom field.
- Repeat for app source, app conversion date and budget source.
- Ask your developer to write app conversions into those fields.
- Go to Dashboard, open your dashboard and select Edit dashboard.
- Select Add widget and add Lead Source Report from General.
- Add a condition on the app source field, then save.
The finished widget lists, by the CRM's own source, every opportunity the app tool also claims. That single table shows how much of the app's credit is new demand and how much is demand the website had already captured.
Overwriting the CRM's source with the app tool's source when the app data arrives. You lose the original record of how the person found you, and it cannot be recovered. Always write the app's view to separate, clearly named fields, and leave the platform's attribution untouched.
Choosing the rule
Three rules make sense for a service business. First known touch wins: whichever system saw the person earliest gets the credit, which favours awareness channels. App wins when the conversion happened in the app: useful when the app is your main booking route. CRM wins unless the CRM source is blank: the simplest, and fair when most enquiries come through the website.
Pick one, write it down and keep it for at least two quarters. Changing the rule monthly makes every trend line meaningless. The marketing attribution overview covers the wider trade-offs between rules.
Suppose 80 clients complete in a quarter. The CRM's first source credits organic search with 32 and the app campaign with 9, while the app tool credits its paid campaigns with 27. Now apply a first known touch rule. If 12 of the app tool's 27 claims came from people the website had already seen, the budget report would credit app campaigns with 15. On those numbers, app spend is still justified, at a lower level than the app tool suggests. All figures are illustrative.
Imagine the app produces 400 valuation requests a quarter from four paid campaigns, and install reports favour the cheapest campaign. Once app source reaches the CRM, say the won-instruction count shows that campaign produced 6 instructions against 21 from the most expensive one. If that held over two quarters, budget would move towards the expensive campaign. The figures are made up.
What it does for a company that has stopped growing
Firms in this position often keep app campaigns running because the app tool's dashboard looks strong, while revenue stays flat. With both views on the contact you can rank app campaigns by won value instead of installs, and stop funding campaigns that bring cheap users who never buy. App leads can also enter the same instant reply and booking follow-up as website leads, so they stop being a slower lane. For the reporting screen itself, see building a marketing attribution dashboard.
Time involved
The fields and the widget take under an hour; the developer work depends on your app. Weekly, someone in marketing should spend fifteen minutes checking that new app leads carry a source and that counts roughly match the app tool's own numbers. Review the budget rule twice a year, not more.
Send only events that show buying intent. Pushing every app event into the CRM produces thousands of meaningless updates and a contact history nobody can read. Behavioural analysis stays in the app tool.
When not to bother
If fewer than one client in ten ever touches the app, the disagreement is too small to matter: use the CRM source alone and review when the app's share grows. If the app is only a portal for existing clients, it has no acquisition role, and credit belongs to whatever brought them in first. The mobile attribution overview covers mobile enquiries that never involve an app.
Troubleshooting
App fields stay empty for app users. The app tool may only send attribution for installs through a tracked link, so organic installs arrive blank. Treat blank as organic in your rule.
The same person appears twice. The app sent a different email or phone format from your forms. Match on email, and store the app user ID on the contact to catch repeats.
Counts differ from the app tool. Some users convert in the app without giving contact details, so they never reach the CRM. Report the gap as a ratio rather than expecting a match.
Frequently asked questions
Which system is right when they disagree?
Neither, completely. Each is right about the touches it saw. Store both, apply one rule for budget and look at both when diagnosing a single channel.
Do I need AppsFlyer to report on app leads?
No. Any tool that can pass campaign data at conversion will do, and a simple "how did you hear about us" field covers small volumes. What matters is that source data reaches the CRM.
Can the platform track installs directly?
No. The platform tracks people once they identify themselves through a form, booking, chat or call. Install tracking belongs in a mobile measurement tool, and the two meet at the moment of conversion.
Should we stop using the platform's own attribution for app users?
No. The platform records the first enquiry and the latest one, which the app tool cannot see. Keep them and add the app view alongside. For revenue credit by salesperson, see sales attribution.