Two critical GA4 and server-side GTM bugs create duplicate events

According to Matteo Zambon, a Google Tag Manager and Analytics expert, two significant technical issues are affecting Google Analytics 4 implementations. The first involves misleading BigQuery event count estimates, while the second centers on a service worker communication bug that duplicates GA4 hits through server-side Google Tag Manager containers.

The BigQuery estimation issue stems from GA4's free version daily limit of 1 million hits for BigQuery exports. Zambon revealed that GA4's dashboard displays unreliable estimates that fail to match actual BigQuery data volumes. "One of our clients got an email warning they were over the limit, but GA4's dashboard showed they were well under it," Zambon stated. His team verified through direct BigQuery analysis that the client had indeed exceeded the daily threshold.

This discrepancy creates operational challenges for organizations managing data export costs and compliance with platform limitations. According to Zambon, businesses should "always verify your event count directly in BigQuery" rather than trusting GA4's internal estimates. The GA4 audit framework emphasizes systematic data quality assessment and direct verification procedures to maintain analytical reliability.

The second issue involves a communication breakdown between Google's gtag service and the service worker component. Giovani Ortolani Barbosa, a digital analytics specialist, documented this duplication bug affecting server-side GTM implementations. According to Barbosa's analysis, "there is a bug in how gtag communicates with the service worker" that prevents proper event acknowledgment.

The service worker implementation began earlier this year when Google tag started leveraging service workers to enhance data collection reliability. The March 4, 2025 official release notes confirmed that "Google tag now uses service workers, when available, to send data to server-side Tag Manager, improving performance and measurement reliability."

However, the current bug disrupts this intended functionality. Barbosa explained that "gtag sends the suppressSuccessCallback = true parameter to the service worker, so the service worker doesn't communicate the result back to gtag in the main page, resulting in the duplicate hit." This technical malfunction affects all events sent when the service worker is loaded and registered on webpages.

The duplication typically excludes events that fire during page load because "the service worker may not be up yet" during initial page rendering. Barbosa's debugging session revealed that older versions of gtag used "suppressSuccessCallback = false" which properly waited for service worker confirmation before proceeding.

Organizations implementing server-side GTM face particular challenges from this bug. According to Barbosa, event numbers for multiple platforms including "GA4, Meta, Google Ads, etc., might be artificially inflated—not because your campaigns are crushing it, but because of a technical glitch."

Current workaround options include blocking specific service worker files or removing iframe elements through JavaScript implementations. Barbosa suggests using "a mix of window load event listener and MutationObserver" to handle iframe removal. However, he advises caution with server-side GTM triggers that automatically abort service worker requests, warning that such solutions could "cause a lot of valid events to be dropped" once Google resolves the underlying issue.

The technical implications extend beyond simple data collection problems. For organizations using automated bidding algorithms that depend on conversion data accuracy, these bugs create cascading effects on campaign optimization and budget allocation decisions. Cost data import functionality through BigQuery, Amazon Redshift, and other sources requires precise event counting to maintain attribution model integrity.

Data analysts using BigQuery for advanced analysis must now implement additional verification steps. The BigQuery schema changes introduced fields like batch_page_id and batch_ordering_id to improve event sequencing analysis, but these enhancements become compromised when duplicate events distort the underlying data structure.

Technical teams managing GA4 implementations should immediately audit their current tracking setups. The verification process includes checking for duplicate events in server-side containers, comparing GA4 dashboard estimates against direct BigQuery queries, and implementing temporary workarounds while awaiting Google's official fixes.

Industry professionals responding to these revelations highlighted broader concerns about measurement accuracy amid increasing privacy restrictions. Some suggested transitioning to server-side tracking alternatives, while others noted the solutions' value for organizations lacking server-side infrastructure capabilities.

The timing of these issues coincides with Google's broader tracking infrastructure changes. Recent updates include Tag Assistant tool unification and enhanced app measurement capabilities, suggesting ongoing platform modernization efforts that may inadvertently introduce temporary instabilities.

Marketing teams should establish monitoring protocols to detect similar issues before they affect campaign performance. This includes regular data validation procedures, cross-platform event count comparisons, and direct BigQuery analysis for organizations approaching the daily export limit.

The service worker duplication bug particularly affects e-commerce implementations where conversion tracking accuracy directly impacts revenue attribution. Online retailers using GA4 for funnel analysis must account for potential event inflation when interpreting user behavior data and conversion rate calculations.

Organizations implementing custom event data import functionality face additional complexity from these bugs. External data integration depends on accurate event matching, and duplicate events could compromise the alignment between GA4 data and external systems like CRM platforms or loyalty programs.

Technical solutions remain limited until Google addresses the underlying service worker communication protocol. Barbosa's GitHub documentation provides code examples for implementing iframe removal workarounds, but organizations must weigh the risks of blocking legitimate service worker functionality against the benefits of preventing duplicate events.

The BigQuery estimation issue affects cost planning for organizations approaching GA4's free tier limits. Companies considering upgrades to Analytics 360 for increased data processing capacity need accurate usage metrics to make informed decisions about platform investments and data architecture requirements.

Timeline

• Early 2024: Google begins implementing service workers for gtag communication to improve data collection reliability
• March 4, 2025: Google officially announces service worker integration for enhanced GTM server-side data transmission
• March 5, 2025: Technical specialists begin publishing verification methods and implementation details for service worker deployments
• June 13, 2024: GA4 introduces Custom Event Data Import functionality for external data integration
• July 16, 2024: Google announces BigQuery schema additions with batch_page_id and event ordering fields
• September 2024: Community reports begin surfacing about duplicate events in server-side GTM implementations
• Recent weeks: Matteo Zambon and Giovani Ortolani Barbosa document specific technical issues affecting GA4 and server-side GTM
• January 2, 2025: Google unifies Tag Assistant tools after user feedback for simplified debugging

Summary

Who: Digital analytics experts Matteo Zambon and Giovani Ortolani Barbosa identified these technical issues affecting organizations using Google Analytics 4 with server-side Google Tag Manager implementations.

What: Two critical bugs impact GA4 data accuracy: misleading BigQuery export count estimates that underreport actual usage, and a service worker communication failure that creates duplicate events in server-side GTM containers.

When: The BigQuery estimation issue affects organizations approaching the 1 million daily hit limit, while the service worker bug emerged following Google's 2024 implementation of service workers for improved data collection reliability.

Where: These problems affect GA4 properties using BigQuery exports and websites implementing server-side Google Tag Manager with active service workers, particularly impacting e-commerce sites and organizations with complex tracking requirements.

Why: The BigQuery bug stems from internal estimation algorithms that fail to match actual data volumes, while the service worker issue results from a communication protocol malfunction where gtag cannot properly confirm event delivery with the service worker component.