The audit examines a GA4 property across 10 thematic categories - from administrative settings to conversion optimization. Here's the map: what each category checks and why it matters. (We deliberately describe scope, not exact decision thresholds - those are part of the audit methodology.)
1. Configuration (administrative settings).
The property's foundations: currency, time zone, data retention. Errors here distort every other report - a wrong time zone corrupts hourly data, a wrong currency falsifies revenue.
2. Implementation.
The measurement code on the site: does the GA4 tag work, avoid duplicating events, and record every visit. This is where the live scan operates.
3. Events & conversions.
Behavior tracking: event configuration and naming (the snake_case convention), key event marking, recording continuity of the most important interactions.
4. E-commerce (stores only).
Sales measurement reliability: completeness of purchase events along the whole path, order values, transaction deduplication, product data quality, discount codes, the purchase funnel.
5. Channels & attribution.
Where traffic really comes from: UTM parameters, channel groups, the attribution model, lookback windows, referral exclusions (the classic: a payment gateway "stealing" conversions).
6. Data quality.
Trustworthiness of the collected numbers: anomalies, bot traffic, duplicated events, tracking gaps.
7. Integrations.
Connections to the Google ecosystem: Google Ads (conversion import!), Search Console, BigQuery - without them GA4 doesn't use its full potential.
8. Limits & technical.
Whether the configuration stays within Google's limits (event counts, parameters, cardinality) - exceeding them means silent data loss.
9. Privacy.
Consent Mode v2 and protection against personal data (PII) leaking into reports. An area of growing legal weight.
10. CRO.
Conversion rates across dimensions: devices, landing pages, channels - where you lose customers, and whether it's a site problem or a measurement problem.
How to read this in the report
Each category shows its checkpoint balance (passing / warnings / errors), and opening a checkpoint reveals the details: what was detected (with evidence), why it matters, how to fix it, and an implementation effort estimate (a standard change vs. a project requiring a developer). The full interactive checkpoint list is available in the app.
Related articles: