Question 7: Self-hosted or hosted privacy-friendly analytics - which is better for a small team, and which products fit each option?

Self-hosted vs. hosted privacy-friendly analytics

For a small team, hosted privacy-friendly analytics is usually the better starting point. You get most of the privacy benefits without taking on server maintenance, upgrades, backups, monitoring, and security responsibilities.

Self-hosting becomes more attractive when you have strict data residency requirements, want complete infrastructure control, or already have someone comfortable maintaining production services.

The important distinction is that privacy-friendly analytics doesn't necessarily mean self-hosted. Several hosted products minimize personal data collection, avoid advertising profiles, and offer cookie-free tracking.

1. Comparing the two approaches

Consideration

Hosted

Self-hosted

Setup

Very easy

Requires infrastructure

Maintenance

Vendor handles it

Your team handles it

Privacy control

Good, vendor-dependent

Greater infrastructure control

Data ownership

Depends on contract and export options

Direct database control

Data residency

Depends on provider

You choose hosting location

Security

Vendor manages platform security

Shared responsibility, largely yours

Scalability

Usually automatic

You manage capacity

Cost

Subscription

Hosting plus engineering time

Small-team suitability

Excellent

Good if technically equipped

An important caveat: Self-hosting does not automatically make analytics more private. You still need sensible retention policies, access controls, secure infrastructure, and appropriate consent management.

2. Best hosted analytics products

These are good choices when you want analytics working without assigning someone to maintain it.

1.

Plausible Analytics

Best overall for small teams

Clean dashboard with visitors, pageviews, referrals, campaigns, and goals.

Cookieless by default.

EU-based infrastructure.

Hosted service includes managed updates and backups.

Also has a self-hosted Community Edition.

Best for: Marketing websites, SaaS landing pages, blogs, and small businesses.

Trade-off: Some advanced features are exclusive to the hosted service.

Website · Documentation

2.

Umami

Best budget-friendly option

Simple, modern analytics dashboard.

No cookies or automatic personal-data collection.

Custom events and website analytics.

Hosted and self-hosted versions.

MIT-licensed open-source software.

Best for: Small development teams that want simplicity, flexibility, and low operating costs.

Trade-off: Less comprehensive than a full Google Analytics replacement.

Website · Documentation

3.

Fathom Analytics

Very simple, polished dashboard.

Cookieless analytics.

Custom events and campaign tracking.

Managed hosting and maintenance.

Best for: Teams that value simplicity and want minimal configuration.

Trade-off: Proprietary hosted product, with no equivalent modern self-hosted edition.

Website

4.

Simple Analytics

Privacy-focused, cookieless tracking.

Minimal dashboard and straightforward reports.

EU-based hosting.

No need to operate analytics infrastructure.

Best for: Teams that want basic website statistics without unnecessary complexity.

Trade-off: Less flexibility and no full self-hosted product.

Website

3. Best self-hosted analytics products

These are appropriate when keeping visitor data on infrastructure you control is important.

1.

Umami

Best overall self-hosted choice

Open-source MIT license.

Relatively straightforward deployment using Docker.

PostgreSQL-backed.

Custom events, dashboards, and website reporting.

No analytics subscription required.

Best for: A small technical team with basic server administration skills.

GitHub repository

2.

Matomo

Best feature-rich option

Full-featured alternative to Google Analytics.

Traffic, events, goals, and ecommerce analytics.

Advanced reporting and optional heatmaps and session recordings.

Open-source On-Premise edition.

Greater control over data storage and retention.

Best for: Teams that need more comprehensive analytics and have someone to maintain the platform.

Trade-off: More complex deployment, maintenance, and privacy configuration.

Website · Documentation

3.

Plausible Community Edition

Open-source AGPL-licensed version.

Lightweight tracking and simple reporting.

Cookieless design.

You control the infrastructure and stored analytics data.

Best for: Teams that want Plausible's philosophy and interface but need to operate their own servers.

Trade-off: Some cloud features aren't included, and releases are less frequent.

GitHub repository

4.

PostHog

Product analytics, event tracking, funnels, and retention.

Session replay, feature flags, and experimentation.

Open-source components and self-hosted deployment.

Hosted cloud option too.

Best for: Software teams that want to understand how customers use an actual application, not just website traffic.

Trade-off: Considerably more operationally demanding than Umami. Self-hosting is resource-intensive.

Website · Self-hosting documentation

5.

GoatCounter

Lightweight, open-source web analytics.

Minimal resource requirements.

Simple traffic and referral reporting.

Self-hostable.

Best for: Small websites, personal projects, and teams that want the smallest possible analytics footprint.

Website · GitHub repository

The strongest distinctions are:

Umami prioritizes simplicity.

Matomo prioritizes reporting depth.

Plausible prioritizes privacy and a clean experience.

PostHog prioritizes product behavior and experimentation.

GoatCounter prioritizes minimalism.

These aren't interchangeable products. In particular, PostHog is more of a product analytics platform than a straightforward website analytics replacement.

4. Quick product comparison

Product

Hosted

Self-hosted

Best use

Plausible

Yes

Yes

Website analytics

Umami

Yes

Yes

Lightweight analytics

Fathom

Yes

No*

Simple reporting

Simple Analytics

Yes

No

Basic website metrics

Matomo

Yes

Yes

Advanced web analytics

PostHog

Yes

Yes

SaaS product analytics

GoatCounter

Yes

Yes

Minimal website analytics

*Fathom Lite exists as an older open-source version, but it isn't equivalent to the current hosted product.

5. What does it cost?

Indicative cost structure, rather than guaranteed quotes:

Product

Hosted cost

Self-hosted cost

Plausible

Paid subscription, entry plans around $9/month

Free software + server

Umami

Free tier available; paid plans available

Free software + server

Matomo

Paid subscription

Free core + server

PostHog

Generous free allowance, then usage-based

Free software + substantial infrastructure

GoatCounter

Free/paid hosted options

Free software + server

Fathom

Paid subscription

Not available as a current equivalent

Remember to calculate the actual cost of self-hosting:

VPS or cloud infrastructure.

Database storage and backups.

Monitoring and security.

Software updates.

Engineering time.

For a small team, saving $10–30 per month isn't necessarily worthwhile if it creates another system someone must maintain.

6. My recommendation by scenario

If you're primarily measuring website traffic

Plausible Cloud

My first choice

Simple, privacy-focused, managed, and easy for non-technical teammates.

If you want to minimize recurring costs

Umami Self-hosted

My first choice

A good balance of functionality, simplicity, and infrastructure control.

If you need detailed analytics

Matomo On-Premise

My first choice

Better when you need comprehensive reporting, ecommerce tracking, and advanced analysis.

If you're building a SaaS application

PostHog Cloud

My first choice

Funnels, retention, feature adoption, session replay, and experimentation are more valuable than basic pageview statistics.

7. Privacy considerations that matter regardless of hosting

Whichever platform you choose, check these things:

Data collection: Does it collect IP addresses, persistent identifiers, device fingerprints, or personally identifiable information?

Cookies: Does it use cookies or other persistent browser storage?

Data residency: Where is visitor data stored and processed?

Data processing agreement: Does the vendor offer a suitable DPA if hosted?

Retention: Can you automatically delete older analytics data?

Access control: Can you restrict access by team member?

Consent requirements: Does your specific implementation require visitor consent?

One important distinction: cookieless tracking doesn't automatically mean consent is unnecessary under every privacy law. The data collected, tracking methods, jurisdiction, and implementation all matter.

8. Final verdict

For a small team, I'd narrow the shortlist to just three:

Plausible Cloud: Best if you want to focus on running your business rather than maintaining infrastructure.

Umami Self-hosted: Best if you have basic technical skills and want ownership without a complicated analytics stack.

Matomo: Best if you need more sophisticated analytics and reporting.

I wouldn't start with PostHog unless you're measuring application behavior, funnels, and product adoption rather than primarily website traffic.

One question that would help narrow this down

Are you primarily looking for analytics for a marketing website, a SaaS/web application, or both?

That distinction changes my recommendation considerably, particularly between Plausible, Umami, Matomo, and PostHog.