Engineering brief
Beating Google Analytics: 3x Cheaper, 97% Data Quality
This engineering brief covers Beating Google Analytics: 3x Cheaper, 97% Data Quality, with practical context for AI and developer-tool decisions.
The Brief
Delivery Hero’s custom tracking system reached 97% order match accuracy, up from Google Analytics’ 85%, cutting costs 3x. That precision came from compile-time validation, but required years of optimization and a doubled pipeline during migration.
Decision relevance
Read this for workflow impact, implementation trade-offs, and the claims that need technical scrutiny before they reach team planning.
Summary
Delivery Hero faced a forced migration from Universal Analytics. Instead of moving to GA4, they built an internal user tracking system to solve real-time data needs, event limits, and GDPR constraints. The initial MVP was a simple API with Pub/Sub and BigQuery, later extended with data validation, archival, and code-generated schemas.
Data quality rose from Google's 85% order match rate to 97% by shifting to compile-time event validation rather than asking developers to fix data manually. Cost per message dropped 25% after one year and eventually reached three times cheaper through archival, storage optimization, and on-demand nodes.
They avoided data loss during migration with a doubled pipeline—sending data via both old and new systems—though this multiplied costs for months until users updated apps. Pod crashes that caused data loss were solved by making requests synchronous, accepting a 7x latency increase.
The big lesson: a focused team with clear KPIs can outperform a generic SaaS product. Full control over priorities, cost, and privacy justified the effort, but it required rigorous load testing, progressive rollouts, and observability that was initially lacking.
Why It Matters
Build vs. buy analytics can yield massive cost savings and data quality improvements if you control your own pipeline.
Editorial analysis
Key claims
- With careful KPIs and iterative optimization, in-house tracking can beat commercial tools on cost and data quality.
Practical use cases
- Use this as input for tooling evaluation, workflow planning, and technical due diligence.
Risks / caveats
- The assumption that Google’s data accuracy is unbeatable.
Who should care
- Engineering managers, tech leads, and CTOs evaluating AI or developer tooling decisions.
Related topics
Bottom Line
With careful KPIs and iterative optimization, in-house tracking can beat commercial tools on cost and data quality.
Watch
This video is blocked due to your privacy settings. To watch this video, please accept YouTube marketing cookies.
Related breakdowns
Rust prevents bugs you didn't know you could eliminate at compile time
Rust's type system prevents more than memory bugs. Learn how ownership, lifetimes, and the type-state pattern eliminate double-use errors, resource leaks…
Document generation as code: from days of debugging to sub-2ms PDFs
Using Rust, Typst, and content-addressable storage, a developer achieved sub-2ms PDF rendering and made document generation versioned and reproducible.
AI Governance Begins with Your Test Suite, Not a New Policy
AI governance is boring: tests, scanning, checklists. Sarah Wells on hidden talent and cost risks.
Get TL;DW
Too Long; Didn't Watch.
A concise breakdowns of the AI and devtools videos that actually matter for engineering leaders.
Free. Weekly. No hype.
Video and thumbnails remain the property of their respective creators. tldw.news provides editorial analysis, commentary, and discovery links to original content.