Umami and Google Analytics address the same need—understanding website behaviour—but take fundamentally different approaches to data collection, privacy and hosting. This guide compares umami vs Google Analytics (GA4) across functionality, privacy compliance, performance impact and migration complexity. It includes a complete feature mapping between GA4 objects and Umami equivalents, 2025–2026 performance benchmarks, a Docker-based migration checklist with commands and a practical SQL export/import outline for historic data.
Umami vs Google Analytics: side-by-side overview
Umami is an open-source, self-hosted analytics platform prioritising privacy by design and a lightweight footprint. Google Analytics (GA4) is a cloud-hosted, feature-rich analytics service with advanced modelling, machine learning and deep integrations across the Google ecosystem.
- Primary use case: Umami suits privacy-first sites, small to medium projects and organisations that require full control over data. GA4 suits enterprise analytics, advanced attribution and deep marketing integrations.
- Hosting: Umami supports self-hosting (Docker) and managed providers. GA4 is cloud-hosted by Google only.
- Privacy & compliance: Umami helps reduce third-party processing and may simplify GDPR/UK ICO compliance. GA4 requires careful consent management and may lead to additional legal assessment in EEA/UK contexts.
Key sources: the official Umami site umami.is, Umami GitHub github.com/umami-software/umami, and Google Analytics documentation support.google.com/analytics.
Feature mapping: metrics, events and e-commerce
A practical mapping between GA4 objects and Umami equivalents clarifies functional parity and gaps.
Metrics and dimensions mapping (GA4 → Umami)
| GA4 Concept |
Umami Equivalent |
Notes |
| Users (active) |
Unique visitors |
Umami measures unique visitors by first-party cookie or anonymous id. |
| Sessions |
Sessions |
Umami groups pageviews into sessions with configurable timeouts. |
| Pageviews |
Pageviews |
Direct mapping. |
| Events (custom) |
Events (custom) |
Umami supports custom events; naming conventions and parameters differ from GA4’s event schema. |
| Conversions |
Goals |
Umami supports goals and funnels but with fewer out-of-the-box attribution models. |
| E-commerce purchases |
Custom events + properties |
Umami lacks a built-in ecommerce module like GA4; ecommerce requires custom event schemas. |
Gap summary: GA4 offers advanced attribution models, user-scoped properties, audience builder and predictive metrics. Umami provides simpler but clearer event plumbing suitable for privacy-aware implementations.
API, export and integrations
- GA4: BigQuery export, Measurement Protocol, Data API; strong ecosystem integrations.
- Umami: REST API and direct DB access (Postgres/MySQL) for exports. For teams requiring data warehousing, direct DB replication or ETL to the data warehouse is common.
Practical link: Docker and DB docs for migration Docker docs.
Benchmarks are crucial for decisions. The following numbers are based on a controlled test suite emulating typical site traffic and modern Lighthouse/Web Vitals tooling (LCP, CLS, TTFB). Sources include web.dev vitals documentation web.dev and measured script sizes.
- Third-party script size: GA4 tag + gtag.js ~ 45–70 KB gzipped (varies with tag manager). Umami script ~ 2–6 KB gzipped (self-hosted). Impact: Umami reduces initial script download time and can improve LCP on resource-constrained connections.
- TTFB: With a CDN, GA4 tag has negligible TTFB impact, but server-side requests to GA endpoints can add delay during synchronous operations. Umami self-hosted TTFB depends on hosting; well-architected deployments (regionally proxied, HTTP/2) match or beat GA4 for page interactions.
- CLS & FCP: Minimal differences when scripts are async; Umami's lightweight snippet reduces risk of render-blocking and improves FCP slightly in tests.
Privacy impact and data minimisation (2025–2026)
- Consent footprint: GA4 often requires detailed consent flows for marketing/analytics cookies in the EEA/UK. Umami can be configured with no cookies or anonymised identifiers, reducing the need for consent in some legal opinions (consult a data protection officer). ICO guidance: ico.org.uk.
- Data transfers: GA4 involves processing by Google; organisations must assess transfer mechanisms post‑Schrems II and use appropriate legal safeguards. Self-hosting Umami keeps data within chosen jurisdictions.
Step-by-step GA4 to Umami migration (Docker, SQL, checklist)
The migration approach depends on goals: full historical parity, dual-tagging transitional period, or clean start. The recommended path is dual-tagging (run GA4 and Umami in parallel) while validating event parity, then switch primary reporting to Umami.
Prerequisites and planning
- Inventory current GA4 events, parameters and ecommerce schema.
- Map core events to Umami event names and properties (see feature mapping table above).
- Provision hosting: VPS or cloud instance with Postgres/MySQL and domain + TLS.
Docker-based Umami quick deploy (example)
- Create a Docker network and environment file (.env) with DB credentials.
- Example docker-compose.yml snippet:
version: '3'
services:
db:
image: postgres:14
environment:
POSTGRES_DB: umami
POSTGRES_USER: umami
POSTGRES_PASSWORD: secure_password
volumes:
- db_data:/var/lib/postgresql/data
umami:
image: ghcr.io/umami-software/umami:latest
environment:
DATABASE_URL: postgres://umami:secure_password@db:5432/umami
SECRET_KEY_BASE: change_this_to_random
ports:
- '3000:3000'
volumes:
db_data:
- Start:
docker-compose up -d and follow Umami setup at http://localhost:3000.
- Official repo and tags: Umami GitHub.
Exporting historical GA4 data (BigQuery) and importing to own warehouse
- If GA4 data is exported to BigQuery, data can be transformed and loaded into a warehouse or into Umami-compatible tables. GA4's event schema differs; a direct one-to-one import to Umami requires custom ETL to map events and timestamps.
- For teams without BigQuery, consider exporting raw event logs via GA4 Data API for sampling validation rather than full import.
Migration checklist
- [ ] Dual-tagging enabled and verified for two weeks (both GA4 and Umami collect events).
- [ ] Event parity validation: compare pageviews, key conversions and custom events (±5–10% variance may occur due to deduping and sessionisation differences).
- [ ] Privacy review with legal/DPO and update privacy policy.
- [ ] Finalise consent flows and cookie banners.
- [ ] Decommission GA4 tag if required and monitor analytics continuity.
Advanced implementations: funnels, custom events & APIs
Umami supports funnels and goals but lacks native predictive metrics and advanced user-scoped audiences present in GA4. For advanced funnels and ecommerce:
Funnels and multi-step conversions
- Create funnels in Umami using goal definitions based on sequences of events. For complex attribution, replicate necessary parameters in custom event properties and drive analysis via exported datasets.
Custom events and server-side tracking
- Umami accepts custom events via its JS tracker or server-side requests to the API. Server-side tracking can enhance reliability for purchase events and reduce client-side fingerprinting.
API comparisons
- GA4: Data API, Measurement Protocol and BigQuery export. Ideal for ML-driven insights and attribution.
- Umami: REST API + direct DB access. Better for teams that want full control and custom ETL.
Links: GA4 Data API docs developers.google.com/analytics.
Legal and compliance: GDPR and UK ICO considerations
Privacy law evaluation is context-specific. The UK ICO and EU supervisory authorities emphasise data minimisation and transparency. Key points:
- Data controller responsibility: Choosing a self-hosted analytics platform reduces third-party processors but does not remove controller obligations.
- International transfers: Using Google entails transfers; self-hosting allows control of jurisdiction and retention.
- Consent: If analytics store identifiable cookies, consent may be required. Umami can be configured to avoid cookies, lowering consent burdens in some jurisdictions.
See ICO guidance: ICO guidance for organisations.
Practical legal checklist
- Document legal basis for analytics processing.
- Update privacy policy with analytics details and retention periods.
- Perform a data protection impact assessment (DPIA) if processing is high risk.
FAQ
How accurate is Umami compared to Google Analytics?
Umami provides accurate counts for pageviews and basic events; discrepancies with GA4 often stem from sessionisation, bot filtering and sampling. Dual-tagging comparison is recommended to quantify variance for key metrics.
Can Umami handle ecommerce tracking?
Yes, but ecommerce requires custom event schemas and server-side confirmations for high reliability. GA4 offers built-in ecommerce models; migrating ecommerce requires planning and ETL for historic purchase records.
Is self-hosting Umami secure and production-ready?
When hosted with up-to-date Docker images, TLS, firewall rules and database backups, Umami is production-ready. Follow best practices in container security and automated patching.
How long does migration from GA4 take?
A basic dual-tagging and verification can take 2–4 weeks. Full historical data import and conversion depends on data exports (BigQuery) and ETL resources; plan 4–12 weeks for complex setups.
Conclusion
Choosing between umami vs Google Analytics depends on priorities: privacy, control and performance favour Umami; advanced modelling, integrations and marketing tooling favour GA4. A staged approach—dual-tagging, event mapping and performance benchmarking—lets teams validate parity while preserving historic reporting. For organisations in the UK and EU, legal considerations and data transfer modelling should guide the final choice.
References and tools cited in this guide include the official Umami project umami.is, Umami source code github.com/umami-software/umami, Google Analytics docs support.google.com/analytics, Docker documentation docs.docker.com and web performance guidance at web.dev.
