Docs

Google Mobile Ads Next-Gen SDK: Ad Revenue Tracking

Google's Mobile Ads Next-Gen SDK is now preferred over the legacy SDK, sunsetting 2028 — what changes in ad-revenue and impression event tracking.

Revenue

Google designated the rewritten Google Mobile Ads (GMA) Next-Gen SDK as the preferred SDK for Android in early July, formally moving the long-standing legacy SDK to deprecated status. Support ends June 2027, full sunset June 2028 — a clock that’s now running for every app monetising through AdMob or Ad Manager. That’s a long runway on paper, but ad-revenue instrumentation is exactly the kind of dependency that gets migrated last and breaks quietly, because impression and revenue events keep flowing right up until a mediation callback or reporting field changes shape underneath you.

The risk isn’t that the migration fails outright — Google has an incentive to make it work — it’s that ad-revenue tracking is rarely built to notice a partial regression. If a mediation adapter reports impressions slightly differently, or a revenue callback fires with a different payload structure, your dashboards can keep showing numbers that look plausible while being subtly wrong. Ad revenue also tends to be the metric with the least redundant verification: unlike a paywall or subscription event, there’s often no independent record to reconcile against beyond the SDK’s own reporting, which makes migration-window discrepancies easy to miss until a monthly revenue reconciliation catches a gap that’s already weeks old.

Data Points to Track

  • Impression and click event volume, compared day-over-day across the migration cutover, since a drop or spike that doesn’t match traffic patterns signals a mediation or callback change rather than a genuine demand shift
  • Revenue-per-impression by ad network and mediation adapter, tracked separately for legacy vs. Next-Gen SDK builds during any staged rollout, to isolate whether a specific network’s reporting changed shape
  • Mediation waterfall fill rate and latency, since the Next-Gen SDK’s rewritten mediation layer can behave differently under the same waterfall configuration
  • SDK version in use per app build, logged as an explicit dimension on every ad event, so revenue anomalies can be attributed to a specific SDK generation rather than blended across a mixed rollout
  • Callback payload schema, diffed between legacy and Next-Gen SDK versions for each integrated ad network, to catch silent field renames or type changes before they corrupt downstream revenue calculations

Setup Steps

  1. Inventory every ad network and mediation adapter currently integrated through the legacy Mobile Ads SDK, since each one migrates to Next-Gen on its own timeline and some may lag.
  2. Stand up parallel tracking for legacy and Next-Gen SDK builds if you’re staging the rollout, tagging every impression, click, and revenue event with the SDK generation that produced it.
  3. Diff callback payloads between the two SDK versions for each network in a test environment before shipping the migration to production, since silent field changes are the most likely source of downstream reporting errors.
  4. Reconcile SDK-reported revenue against ad-network dashboards during the migration window specifically, not just as a routine monthly check, since this is when discrepancies are most likely to appear.
  5. Set an explicit migration deadline against your own release calendar, well ahead of the June 2027 support cutoff, so the switch happens on your team’s schedule rather than as a forced late migration under time pressure.

Actionable Insights

A revenue-per-impression figure that shifts for one ad network but not others during a mediation SDK migration almost always points to that network’s specific adapter, not a platform-wide issue — narrow the investigation there first. If impression volume holds steady but revenue drops, the mediation waterfall or fill rate is the more likely culprit than the event tracking itself. And treating the migration window as a distinct, tracked period — rather than a single flip you check once — is what catches a slow-building discrepancy before it shows up as an unexplained gap in a monthly revenue report.

Expert help

Need help tracking this in your app?

Our team sets up analytics pipelines for mobile and web teams every day. Talk to us and get your first events flowing in under an hour.

Talk to an expert