Skip to main content
Extole → AppsFlyerIntegration guide

Connecting Extole to AppsFlyer

Extole runs enterprise advocacy programmes where referral, review and social sharing all produce rewards, which makes several different behaviours compete for credit on the same order. AppsFlyer exists because mobile app attribution is a different problem from web attribution, and on iOS a large share of it is now aggregated and delayed by design rather than by accident. The join between them succeeds or fails on one thing: whether the AppsFlyer device identifier, appsflyer_id reaches the Extole advocacy event before the conversion happens, because AppsFlyer can only match a conversion it can tie back to a click it recognizes.

Free30 minutesNo obligation

Specific to this pair

Extole to AppsFlyer: a referred friend needs a device, not an email

Generate Extole share links as AppsFlyer OneLinks, or wrap them, so AppsFlyer's user-invite attribution credits referred installs. Do not resend a friend's first purchase if the SDK already logs it; instead send a distinct referral-qualified event with no revenue, keyed to the stored appsflyer_id. Keep web-only referrals out, and compare referral installs with Extole sign-ups weekly.

AppsFlyer can attribute an install to a referral only if the referral reached the device. Extole knows the friend by the share link they clicked and later by email at sign-up; AppsFlyer knows them by appsflyer_id. If the advocate's share link is a plain web URL, the friend who installs from it arrives in AppsFlyer as organic, or as whichever paid network last showed them an ad, and then Extole and the network both claim the install.

Check whether your Extole share links can be generated as AppsFlyer OneLinks or wrapped in one. AppsFlyer offers user-invite attribution for this case, so the install is credited to the referral as its own source. Do not resend the friend's first purchase from your server if the SDK already logs it; that doubles revenue in AppsFlyer and in every network it forwards to. If you need the referral outcome in AppsFlyer, send a distinct referral-qualified event with no revenue, keyed to the appsflyer_id you stored at sign-up.

Web-only referrals should not go to AppsFlyer at all, since a friend who bought on the website has no install behind the order. The precedence rule then splits by surface: for app users, AppsFlyer's referral source decides; for web orders, Extole's record does. Compare AppsFlyer's referral-attributed installs against Extole's app-originated referred sign-ups each week. A large gap means the link is not surviving the trip through the store.

Extole → AppsFlyer · what breaks

Why this join usually fails — Extole → AppsFlyer

  1. 01

    The AppsFlyer device identifier, appsflyer_id never reaches Extole

    Several advocacy channels can each attach their own parameter to the same visit.

  2. 02

    Three programmes claiming one order

    Referral, advocacy and paid media can all show the same order as theirs. Summing programme-reported conversions across an enterprise advocacy stack overstates total acquisition substantially.

  3. 03

    Web conversions sent to a mobile attribution platform

    This is the common error and it produces confident nonsense. A purchase made on your website has no device identifier and no install behind it. AppsFlyer will either discard it or attribute it to nothing, and the revenue disappears somewhere between the two systems without anyone noticing it went.

  4. 04

    Nothing reconciles afterwards

    AppsFlyer shows the event on the raw data export with its attributed media source. Check the raw export, not the aggregate dashboard, and confirm the event carries a device identifier rather than landing as organic.

Extole → AppsFlyer · the fix

What working looks like — Extole → AppsFlyer

  • The AppsFlyer device identifier, appsflyer_id on every referral

  • Conversions arriving in AppsFlyer

    The event is attributed to the install source, forwarded to whichever networks are configured, and appears in the cohort reporting.

  • Reconciled against the source

Extole → AppsFlyer · architecture

How Extole and AppsFlyer actually connect

  1. 1

    Extole holds the truth

    An order can be attributable to a referral, a share and a paid click simultaneously. Enterprise advocacy platforms make that overlap common rather than rare.

  2. 2

    AppsFlyer needs a click to match

    AppsFlyer needs the device identifier and an event name it recognizes, plus the revenue value for in-app purchases. A server event about a user with no linked device cannot be attributed to an install.

  3. 3

    The identifier is the bridge

    The AppsFlyer device identifier, appsflyer_id, captured at the visit and carried into the Extole advocacy event.

  4. 4

    The event source decides reliability

    Send orders net of reward cost, attributed under a written precedence rule.

  5. 5

    Deduplication at the destination

    AppsFlyer deduplicates in-app events by its own event identifier. Sending the same purchase from both the SDK and your server produces two events and double-counts revenue in every dashboard downstream, including the ad networks AppsFlyer forwards to.

Extole → AppsFlyer · normal or broken

Working vs. broken, side by side — Extole → AppsFlyer

Normal, not a fault

  • The AppsFlyer device identifier, appsflyer_id present on the referral

  • Sent from your order system with advocacy attribution

  • Deduplicated correctly

    AppsFlyer deduplicates in-app events by its own event identifier. Sending the same purchase from both the SDK and your server produces two events and double-counts revenue in every dashboard downstream, including the ad networks AppsFlyer forwards to.

  • Reconciles with the source

Actually broken

  • The AppsFlyer device identifier, appsflyer_id missing on most records

    Several advocacy channels can each attach their own parameter to the same visit.

  • Sent from the browser

    Referral, advocacy and paid media can all show the same order as theirs. Summing programme-reported conversions across an enterprise advocacy stack overstates total acquisition substantially.

  • Counted twice, or not at all

    This is the common error and it produces confident nonsense. A purchase made on your website has no device identifier and no install behind it. AppsFlyer will either discard it or attribute it to nothing, and the revenue disappears somewhere between the two systems without anyone noticing it went.

  • Nobody has reconciled it

Extole → AppsFlyer · diagnose

Diagnostic order — Extole → AppsFlyer

  1. 01

    Count referrals carrying the AppsFlyer device identifier, appsflyer_id

  2. 02

    Follow one real conversion end to end

  3. 03

    Count orders claimed by more than one programme

    The overlap is the overstatement, and at enterprise scale it is rarely small.

  4. 04

    Check what share of your revenue happens in the app at all

    If most purchases are on the web, AppsFlyer is not your conversion destination and the work belongs somewhere else entirely.

  5. 05

    Read the match rate, not the success message

    AppsFlyer shows the event on the raw data export with its attributed media source. Check the raw export, not the aggregate dashboard, and confirm the event carries a device identifier rather than landing as organic.

Extole → AppsFlyer · prove

How it gets proven — Extole → AppsFlyer

Source of truth

Orders flagged by which advocacy programme produced them, with reward costs attached.

  • The AppsFlyer device identifier, appsflyer_id confirmed present on a live referral.
  • In-app revenue events reconciled per transaction against your billing system, with the organic share stated rather than assumed away.
  • Orders reconciled once each under a documented precedence rule, with the overlap measured and stated.

Extole → AppsFlyer · free audit

What we check in the free audit

  • Count referrals carrying the AppsFlyer device identifier, appsflyer_id
  • Follow one real conversion end to end
  • Count orders claimed by more than one programme
  • Check what share of your revenue happens in the app at all
  • Read the match rate, not the success message

Checked against

Orders flagged by which advocacy programme produced them, with reward costs attached.

Get my free conversion tracking audit

Free30 minutesNo obligation

Extole → AppsFlyer · when not to buy

When you don't need this — Extole → AppsFlyer

  • Measure identifier coverage yourself first — pull your last hundred referrals and count how many carry the AppsFlyer device identifier, appsflyer_id.
  • A single referral programme has no competing claimants and needs no precedence rule.
  • If you have no mobile app, this is not a destination for you. If you have one but sell mostly on the web, send web conversions to your ad platforms directly and use AppsFlyer only for install and in-app events.

Extole → AppsFlyer · what we deliver

What gets built — Extole → AppsFlyer

  • Identifier capture into Extole

    Record every parameter present with a timestamp and apply a documented precedence rule rather than letting the last one win.

  • Event source: your order system with advocacy attribution

    Send orders net of reward cost, attributed under a written precedence rule.

  • Dispatch to AppsFlyer

    The AppsFlyer server-to-server events API, posted from your billing system with the device identifier and revenue.

  • Reversals where they exist

    Refunds reverse and reclaim rewards.

Written guarantee

95%+ tracking accuracy

Measured against Orders flagged by which advocacy programme produced them, with reward costs attached.. Miss it and we keep working at no cost.

Get my free conversion tracking audit

From $400 · See packages

FAQ

Extole → AppsFlyer questions, answered

The share link has to carry the referral through the install, typically by building it on an AppsFlyer OneLink so the install is credited to the referral source. A plain web link makes the install look organic.

Free · 30 min

Get Extole and AppsFlyer joined properly

Free30 minutesNo obligation