▸Example Matomo Mobile call (the "before")
import MatomoTracker
let tracker = MatomoTracker(siteId: "1", baseURL: URL(string: "https://your-matomo.example.com/matomo.php")!)
tracker.userId = userId
let event = Event(
tracker: tracker,
action: ["purchase"],
eventCategory: "ecommerce",
eventAction: "purchase",
eventName: "paywall",
eventValue: NSNumber(value: price).floatValue
)
tracker.track(event)
App-binary size affects download conversion (especially on lower-end devices and cellular networks) and storage pressure for users with full phones. Heavy analytics SDKs (Firebase Analytics, AppsFlyer, Branch) routinely add several MB to the binary; Respectlytics's per-platform SDK is under 100 KB.
☑Remove Matomo Mobile cleanly
-
1
Remove the Matomo SDK from your build (
MatomoTracker/org.matomo.sdk:tracker/matomo_tracker) -
2
Remove
Tracker.builder().build()initialisation andtrack(...)call sites -
3
Decide whether your Matomo server stays online (for web traffic) or gets decommissioned
-
4
If you used Matomo's visitor segments alongside web data, plan how you'll bridge web ⇄ mobile analytics under Respectlytics (web isn't Respectlytics's primary surface)
⇋Matomo Mobile vs Respectlytics — smaller app binary
| Matomo Mobile | Respectlytics | |
|---|---|---|
| Typical contribution to IPA size | — see tool note above | < 100 KB |
| Typical contribution to APK / AAB size | — see tool note above | < 100 KB |
| Transitive dependencies | — see tool note above | None (zero third-party) |
❓Frequently asked questions
How much do these SDKs typically add to my binary?
Varies by SDK. Rough order of magnitude on Android (release AAB): Firebase Analytics ~3-4 MB (with transitive Google Play Services), AppsFlyer ~1-2 MB, Branch ~1.5 MB, Mixpanel ~1 MB. Our [SDK Bundle-Size Comparator](/tools/sdk-bundle-size-comparator/) has up-to-date numbers per SDK.
Does binary size actually affect downloads?
Yes — particularly past Apple's 200 MB cellular-download cap and on lower-end Android devices with limited storage. Google's Play Console has internal metrics showing measurable install-rate differences when AAB size crosses common thresholds (50 MB, 100 MB).
How does the size reduction work in practice?
Removing a single ~3 MB SDK from your release build shrinks the binary by roughly that amount. Multiplied by user installs, the cellular-data savings compound. iOS App Thinning handles per-device variants; Android AAB delivers size-optimised installs by ABI / density / language.
Does the smaller SDK come at a feature cost?
Respectlytics intentionally doesn't bundle features it considers separate concerns: no ads, no attribution, no in-app messaging, no A/B testing, no crash reporting. Each of those is a dedicated tool's job. The smaller surface is the deliberate design — you add only what you actually use.