Project Title

GA4 Export Limit Mitigation Strategy

Introduction

In early 2023, I led a critical initiative to future-proof RHS’s digital analytics infrastructure. As our GA4 event volume began exceeding BigQuery’s daily export limit, we faced a growing risk to data continuity and reporting accuracy. I spearheaded the investigation, forecasting, and vendor evaluation process—balancing cost, scalability, and operational impact. This case study outlines how we navigated technical constraints, engaged cross-functional stakeholders, and ultimately aligned on a solution that supports both immediate needs and long-term strategy.

The Challenge

As RHS’s digital footprint grew, we began exceeding the 1 million daily hit limit for GA4’s BigQuery export. This posed a serious risk to our analytics continuity—especially during peak traffic periods—threatening the integrity of our reporting and data-driven decision-making.

My Role

As Product Owner for Optimisation, I collaborated with stakeholders across Digital, Data & Analytics, and IT to assess the impact, forecast future demand, and evaluate scalable solutions. I led the analysis of traffic trends, facilitated vendor demos, and helped shape the recommendation strategy presented to senior leadership.

Tools & Setup

• Google BigQuery: Primary data warehouse for GA4 event data
• GA4 + Server-GTM: Source of digital analytics data
• Snowplow (demoed): Evaluated as a scalable, flexible alternative
• Forecasting Models: Used historical UA data to predict GA4 event volumes

What We Tested

We explored multiple solutions to mitigate the hit limit issue:

• Upgrading to GA4 360 for higher export limits
• Implementing Snowplow as a parallel data collection system
• Using streaming exports only (with limitations)
• Filtering events in BigQuery
• Rebuilding reporting directly in GA4

What I Found

• Strong correlation between 2022 and 2023 traffic patterns validated our forecasting model
• Top 6 events accounted for 95% of hits—making filtering unviable without losing critical insights
• Streaming exports lacked session-level data, limiting campaign attribution
• Snowplow emerged as a cost-effective, scalable, and flexible solution with strong Azure integration

What I Learned

• What worked well: Our cross-functional evaluation process helped us weigh cost, scalability, and operational complexity. The forecasting model and stakeholder engagement were key to building consensus.
• What we decided: After leading a thorough evaluation of Snowplow and GA360, I recommended GA360 as the preferred solution. While Snowplow offered a lower license cost, GA360 provided a faster, lower-friction implementation path—thanks to our existing Google stack setup. The decision was further supported by strategic input from our Director of IT, who emphasized the importance of:
  • Operational simplicity: GA360 aligned with our IT strategy to reduce platform complexity.
  • Speed to implementation: It offered a faster path to restoring full data continuity.
  • Total Cost of Ownership: Factoring in not just license cost, but long-term maintenance and skills required.
• What I’d do differently: Factor in internal platform strategy and operational ownership earlier in the vendor evaluation process.
• Surprises: The manual workaround for session-level data (via BigQuery scripting) highlighted how critical this data is for campaign analysis—strengthening the case for a robust, integrated solution.

📸 Visuals

Forecasting correlation chart (2022 vs 2023 traffic)

This chart illustrates a strong positive correlation between RHS website pageviews in January–February 2023 and the same period in 2022, suggesting consistent seasonal traffic patterns year over year.

GA4 Daily Events Forecast: Jan 2023 Onward

This chart compares actual and forecasted daily GA4 event volumes from January 1 to mid-January 2023, projecting future traffic trends. The forecast anticipates a significant spike in early June, highlighting the need for scalable analytics infrastructure.

Proposed Snowplow Data Architecture for RHS

This diagram outlines the end-to-end data flow and governance model proposed by Snowplow for RHS. It illustrates how data would be created, enriched, modelled, and reported using a resilient, schema-driven pipeline—integrating sources like the RHS website, mobile apps, and server-side events into a Snowplow-managed GCP environment, with outputs feeding into BI tools, CRM, and marketing platforms.

Cost-benefit comparison table

Tool	Cost (Range)	Pros	Cons
GA4 360	High (approx. 2x Snowplow)	Seamless upgrade from GA4 & BigQuery Instantly solves export limit issue Access to Google stack tools (Audience Centre, Attribution, Optimize)	Significantly more expensive than alternatives
Snowplow	Mid-range (charity discount applied)	Fast implementation Azure integrations Customisable data modelling and sessionisation Lower license cost	Requires more setup and internal ownership
Fivetran	Variable (£1k–£6k/month)	SEMrush integration Pay-per-row pricing model	Unclear pricing structure Outdated connectors Complex setup Mixed reviews on support

Feedback Welcome

If you’ve tackled similar challenges around scaling analytics infrastructure, managing GA4 limitations, or evaluating tools like Snowplow or GA360, I’d love to hear your perspective. Whether it’s lessons from navigating platform trade-offs, aligning with IT strategy, or forecasting data growth, feel free to connect — I’m always keen to exchange ideas and explore better ways to deliver value through data-informed product decisions.