GA4 shows you installs.
Airbridge shows you subscribers.
Your GA4 dashboard shows installs. Your RevenueCat dashboard shows subscribers. Nothing joins the two, so you can't see which channel's installs became paying subscribers, or which ones churn before the first renewal. Airbridge joins them, across Meta, Google, TikTok, and Apple Search Ads.
Cross-channel performance is the hard part.
You're running Meta, Google, TikTok, and Apple Search Ads at once, and each one grades its own conversions. The numbers overlap and nothing reconciles them. GA4 attributes installs natively for Google Ads and leans on modeled data everywhere else, so the rest land in "Direct" or "(not set)". The channels you're actively scaling end up the hardest to check. That isn't a setting you can fix. It's what a deduplicated, third-party view is for.
GA4 vs Airbridge, feature by feature
Anything labeled "Growth Plan" needs the Growth Plan. Everything else is included in Core, from $40+/mo.
| Feature | Airbridge | GA4 |
|---|---|---|
| Website traffic & on-site behavior | Not what it's for | The default, and rightly so |
| Conversion postbacks to ad networks | Meta, Google, Apple, TikTok, and 350+ more | |
| Non-Google channel attribution | Deterministic, network by network | Falls into "Direct" or "(not set)" |
| Custom attribution windows | Set per channel and per event | |
| Deep linking & web-to-app | Deferred deep links, included | |
| Trial → paid, by channel | Install → trial → subscribe → renew | |
| RevenueCat / Adapty sync | Native server-to-server | |
| Delayed revenue (renewals) | Attribution persists for the whole lifecycle | Modeled attribution reassigns or drops it |
| Re-attribution vs. acquisition | Separate re-attribution windows | Re-engagement overwrites the original source |
| Fraud Validation Rules | Growth Plan | |
| Touchpoints Overlap Report & Predictive LTV | Growth Plan | |
| Data ownership | Your data, your warehouse | Google-controlled modeling |
Seven things GA4 was never built to do
Plus one it does better than we do.
Ad networks can't learn from data they never receive
Meta, Google, and TikTok optimize by feeding conversions back into their own models. That return trip is a postback, and GA4 doesn't send one. Self-attributing networks partly cover for themselves. Every other channel is flying on install counts alone. Airbridge sends conversion postbacks to Meta, Google, Apple, TikTok, and 350+ more, so the algorithm trains on subscribers instead of installs.
Learn moreA 7-day trial doesn't fit inside a session
GA4's attribution assumes the conversion happens near the click. Subscription apps break that assumption on purpose: the trial converts on day 7, the renewal lands in month two. Airbridge lets you set the attribution window per channel and per event. For web-to-app journeys, Protected Attribution Window (PAW) excludes landing-page CTA touchpoints from attribution, so the original ad keeps credit through the App Store detour. That detour is exactly where a UTM chain snaps.
Learn moreHalf your budget shows up as "Direct"
GA4 attributes app installs natively for Google Ads. For Meta, TikTok, and Apple Search Ads it leans on UTM parameters, which don't survive the App Store redirect. So the channels you're pushing hardest are the ones GA4 sees least. Airbridge integrates with each network directly, so every channel is measured the same way.
Learn moreRe-engagement quietly overwrites where the user came from
When a lapsed user returns through a retargeting ad, GA4 has no concept of a re-attribution window. The return visit takes the credit and the original acquisition source disappears. Airbridge separates the two, so you can define a lapsed audience (30 days inactive, 60 days inactive), push it back to Meta or Google as a retargeting or exclusion list, and still know which channel brought that user in the first place.
Learn moreYour subscriber revenue lives where GA4 can't reach
Trials start in your app. Conversions, renewals, and cancellations happen on Apple and Google billing servers, and RevenueCat or Adapty is what hears about them. GA4 has no server-to-server connection to any of it. Airbridge does, so a renewal in month four still points back to the ad that started it.
Learn moreFraud inflates the channel you're about to scale
GA4 doesn't screen for ad fraud, so click flooding, SDK spoofing, and device-farm installs arrive in your reports as real users. The channel serving them then looks like your cheapest source of growth. Airbridge's Fraud Validation Rules catch click injection, install hijacking, click flooding, SDK spoofing, device farms, and device-ID reset fraud before any of it reaches your reports. Included in the Growth Plan.
Learn moreA good model still needs the whole path
GA4 offers last-click and data-driven attribution, but either model can only weigh the touchpoints GA4 actually recorded. On mobile that stream is missing the non-Google installs entirely, and neither model counts view-through conversions. Airbridge's Touchpoints Overlap Report shows where channels genuinely shared the path to conversion, and Predictive LTV forecasts 180 days out from three days of data, so you can compare channels on what they will be worth rather than only on what they closed. Included in the Growth Plan.
Learn moreThis isn't a migration.
Nothing gets ripped out and nothing gets rebuilt. GA4 goes on doing what it does on your website. Airbridge sits underneath as the system of record for acquisition and revenue, which is the number you actually set budget against. One SDK, no data team, and your existing reporting keeps working.
Learn more"I spent $3k across Meta and TikTok last month.
How many paying subscribers did I get?"
| Step | Airbridge | GA4 |
|---|---|---|
| Where you start | Airbridge revenue report | GA4 install report |
| What you see | Subscribers and revenue, by channel and campaign | 3,000 installs, Google Ads on top |
| Split it by channel | Trial starts, trial-to-paid rate, and renewals per channel | Meta and TikTok show as "Direct" |
| Cross-check the money | RevenueCat is already synced, so the numbers match | Open RevenueCat separately, with no shared key to join on |
| Feed it back to the networks | Postbacks fire automatically, so the algorithm learns | Nothing to send |
| What you do next | Cut the channel with cheap installs and low trial-to-paid, scale the other | Guess, or fall back to CPI |
Priced for the stage you're actually at
GA4 is free, and that's a fair thing to weigh. Here's what adding an MMP actually costs.
Start with a 30-day free trial
See it work on your own campaigns before you pay. You don't talk to sales and you don't sign an annual contract.
Transparent, usage-based pricing
$40+/mo including 500K data points, then $0.0001 per additional data point. Attribution and tracking links are unlimited and never count as data points.
Live in about an hour
Five minutes for the SDK, one-click toggles for Meta, Google, Apple, and TikTok, and RevenueCat or Adapty syncs on its own. You won't need to design an event taxonomy or hire a data engineer first.
Every channel.
Every trial.
Every renewal.
One number.
GA4 tells you what users did. Airbridge tells you which ads paid for them.