Overview
Two Open Source Philosophies
Same Goal, Different Approach
Staminads: Built for pro marketers who need to detect which dimensions drive engaged traffic. Median metrics (TimeScore), advanced traffic categorization with backfill, and full ad network detection.
Umami: Simple analytics with basic funnel reports (same browser/month only). Average-based metrics with device fingerprinting for visitor counting.
Both Fully Open Source
Staminads: AGPL-3.0 license. Full-featured self-hosting with Docker and ClickHouse.
Umami: MIT license. Self-host with Docker, Vercel, or various cloud providers. PostgreSQL or MySQL.
What They Share
Both platforms are built on privacy-first principles:
- Cookieless: No cookies, no consent banners required
- Open Source: Full source code available
- Self-hostable: Run on your own infrastructure
- Full API: REST API for data access
- Team support: Multi-user with role-based access
Where They Differ
The key differences come down to depth of analysis and tracking approach:
Staminads Advantages
- • Median metrics (not averages)
- • Traffic categorization with backfill
- • Full ad network detection
- • No cross-session tracking
- • ClickHouse for fast queries
- • Custom Explore for deep analysis
Umami Advantages
- • Basic funnel analysis (same browser/month)
- • Journey visualization (same browser/month)
- • Retention reports (same browser/month)
- • MIT license (more permissive)
- • Run free on Vercel + Supabase
- • 34K+ GitHub stars community
Umami is a popular open-source analytics platform with 34K+ GitHub stars. It fingerprints visitors by hashing IP + User-Agent with monthly rotating salts, and offers basic funnel analysis, journey reports, and retention tracking — limited to same browser, same month due to how fingerprinting works. MIT licensed and can run for free on Vercel.
Staminads is also open source and self-hostable, but focuses on deeper engagement analysis with median metrics (TimeScore), advanced traffic categorization with historical backfill, and full ad network detection for pro marketers.
Quick Comparison
| Staminads | Umami | |
|---|---|---|
| Tracking Method | Cookieless (session-only, no fingerprinting) | Cookieless (IP+UA hash fingerprinting) |
| Consent Required | No consent needed | Legally contested (ePrivacy Directive) |
| Engagement Metrics | Median values (TimeScore) | Average only (excludes single-page visits) |
| Traffic Categorization | Advanced traffic categorization for pro marketers | None (dashboard filters only) |
| Traffic Backfill | Recompute & modify historical data | None |
| Ad Network Detection | Google, Meta, TikTok, LinkedIn, Microsoft | UTM parameters only |
| Funnel Analysis | Coming soon | Basic (same browser/month only) |
| API Access | Full access (unlimited) | Full access (50 calls/15 sec limit) |
| Pricing | Free (AGPL-3.0, self-hosted) | Free (MIT License, self-hosted) |
| User Identification | Session-only (no cross-session) | Distinct IDs available (cross-session) |
Tracking & Privacy
Both platforms are cookieless, but use different approaches to visitor identification.
The Fingerprinting Question
Umami uses device fingerprinting — hashing IP address + User-Agent + hostname with rotating salts to identify visitors. While Umami claims GDPR compliance, the EDPB Guidelines 2/2023 (finalized October 2024) raise concerns:
- ePrivacy Directive Article 5(3): Fingerprinting explicitly falls within scope and requires consent unless a narrow exemption applies
- Hashing ≠ Anonymization: Under GDPR, hashed data is pseudonymized, not anonymous — IP addresses remain personal data
- Distinct IDs: Optional cross-session tracking moves further toward user identification
Staminads takes a different approach: IP is processed only for GeoIP location (with customizable settings), not for fingerprinting or identification. No device hashing, no cross-session linking. Each session is independent — eliminating the legal gray area.
Privacy & Tracking Comparison
| Feature | Staminads | Umami |
|---|---|---|
| Cookies | No cookies used | No cookies used |
| Device Fingerprinting | No fingerprinting | Yes (IP + UA hash with rotating salt) |
| ePrivacy Directive Compliance | Clear (no device access/fingerprinting) | Legally contested (EDPB Guidelines 2/2023) |
| Visitor Identification | None (session-only) | Hash with monthly rotating salt |
| Cross-session Tracking | Optional (Distinct IDs feature) | |
| IP Address Processing | GeoIP only (customizable) | Hashed for visitor fingerprinting |
| Data Residency | Your servers (you control) | Your servers or Umami Cloud |
Umami's Distinct IDs
Umami offers an optional umami.identify() feature that allows cross-session user tracking. This can be useful for logged-in users but moves away from pure privacy-first analytics. Staminads deliberately doesn't offer this — each session is independent with no way to link users across visits.
Staminads: Session-Only
Staminads processes IP addresses only for GeoIP location (with customizable settings) — not for identification. No hashing, no cross-session linking. Pure session analytics with the clearest privacy stance.
Traffic Categorization
This is where Staminads shines. Advanced traffic categorization with backfill means you can map your marketing mix and recategorize historical data at any time.
Traffic Categorization Features
| Feature | Staminads | Umami |
|---|---|---|
| Channel Groupings | Custom rules to map your marketing mix | None (no channel groupings) |
| Traffic Rules | Set value, clear value, set default | None |
| Custom Dimensions | 10 dimensions you can modify via rules | Event properties only |
| Backfill on Rule Change | Yes (recompute all historical data) | No |
| Dashboard Filters | Yes (all dimensions) | Yes (Is, Is not operators) |
| Ad Network Detection | Google, Meta, TikTok, LinkedIn, Microsoft | UTM parameters only |
| UTM Parameters | Full support + custom | source, medium, campaign, content, term |
| Rule Operators | equals, contains, regex, is empty | None (no categorization rules) |
Ad Network Click ID Detection
Staminads automatically detects click IDs from all major ad networks:
gclid Google Adsfbclid Meta/Facebookttclid TikTokmsclkid Microsoftli_fat_id LinkedIndclid Google DV360Umami relies on UTM parameters only. For ad network traffic without UTMs, you'd need to manually tag every campaign.
The Backfill Difference
In Staminads, when you update a traffic rule (e.g., split "social" into "twitter" and "linkedin"), all historical data is recomputed with the new categorization. Umami has no traffic categorization rules — only dashboard filters for querying existing data. You can't group sources into custom channels or recategorize historical traffic.
Real-time Data Exploration
Both offer dashboards, but database choice matters at scale. Umami uses PostgreSQL/MySQL, which slow down significantly as your data grows.
Real-time Capabilities
| Feature | Staminads | Umami |
|---|---|---|
| Real-time Dashboard | Slows down at scale | |
| Data Availability | Immediate | Immediate (small datasets) |
| Current Visitors | Instant | Last 30 minutes |
| Database | ClickHouse | PostgreSQL or MySQL |
| Data Sampling | Never (full dataset) | Never |
| Historical Query Speed | Sub-second for any range | Slows down as data grows |
Why Umami Slows Down at Scale
Umami uses PostgreSQL or MySQL — databases designed for apps and websites, not analytics. When you ask "how many visitors from France last month?", these databases must scan through every single record to find and count the matches. With millions of pageviews, this takes seconds or even minutes.
Staminads uses ClickHouse — a database built specifically for analytics. It stores data in a way that makes aggregation queries instant, even on billions of events. For high-traffic sites, this is the difference between a responsive dashboard and one that times out.
Deep Data Exploration
Staminads offers a filterable dashboard plus a Custom Explore for deep data analysis. Drill down by any dimension — source, campaign, country, device, or custom dimensions. Umami offers an Insights report with segmentation capabilities.
Engagement Metrics
This is a key differentiator. Staminads uses median values for engagement metrics. Umami uses averages and excludes single-page visits from duration calculations.
Engagement Metrics Comparison
| Feature | Staminads | Umami |
|---|---|---|
| Session Duration Calculation | Median (TimeScore) | Average (excludes bounces) |
| Single-page Session Duration | Measured via focus tracking | Excluded (no second event) |
| Scroll Depth | 25%, 50%, 75%, 100% thresholds | Not built-in (custom events only) |
| Time on Page | Median with focus tracking | Average (time between events) |
| Bounce Rate | Configurable time threshold (default 10s) | Single-event visits / total visits |
| Engagement Score | TimeScore (composite median) | No composite metric |
Why Median Matters
Umami's average can be misleading. Staminads' median shows you typical user behavior.
Single-Page Visit Problem
Umami calculates session duration as "time between first and last event." This means single-page visitors (bounces) are excluded entirely from duration calculations — a significant blind spot. Staminads uses focus tracking to measure engagement on every page, including bounces.
Goals & Conversions
Both platforms support goal and conversion tracking. Umami has funnel and journey reports, but they're limited to same-browser, same-month tracking due to fingerprinting constraints.
Goals & Conversions Features
| Feature | Staminads | Umami |
|---|---|---|
| Goals Dashboard | Dedicated view with KPIs & charts | Goals report |
| Conversion Tracking | ||
| Revenue Tracking | Yes (with median value) | Yes (Revenue report) |
| Funnel Analysis | Coming soon | Limited (same browser, same month) |
| Journey Visualization | Coming soon | Limited (same browser, same month) |
| Retention Reports | Coming soon | Limited (same browser, same month) |
| Goal Breakdowns | By any dimension | By URLs and events |
| Event Properties | 10 custom dimensions | Custom properties |
Umami's Funnel Limitations
Umami offers funnel analysis with 3-7 configurable steps, journey visualization, and retention reports. However, because Umami uses device fingerprinting (IP + User-Agent hash), funnels have significant limitations:
- • Monthly reset: Salt rotates monthly — a visitor on January 30 who converts on February 2 appears as two different people
- • Browser-specific: Each browser has a different fingerprint — switch from Chrome to Safari and you're a new visitor
- • No cross-device: Desktop and mobile are tracked as separate visitors
- • Webviews break tracking: Traffic from Instagram, Facebook, or TikTok webviews can't be linked back to the original user
Staminads focuses on traffic categorization and engagement metrics — funnels are on the roadmap.
Find Your Winning Dimensions
Staminads' goals dashboard reveals which traffic sources, campaigns, and dimensions drive the most conversions. Identify winners by any dimension — source, campaign, country, device, or custom dimensions — all in one view.
Multi-tenant & Teams
Both platforms support multiple sites and team collaboration.
Multi-tenant Features
| Feature | Staminads | Umami |
|---|---|---|
| Multiple Websites | Unlimited | Unlimited |
| Team Support | Yes (workspaces) | Yes (teams with access codes) |
| Website Transfer | Between workspaces | Between personal and teams |
| Data Retention | Unlimited (self-hosted) | Unlimited (self-hosted) |
User Roles & Permissions
User Management
| Feature | Staminads | Umami |
|---|---|---|
| Account Roles | Owner, Admin, Editor, Viewer | Administrator, User, View Only |
| Team Roles | Per-workspace permissions | Team Owner, Manager, Member, View Only |
| Granular Permissions | Per-workspace, per-site | Per-team, per-website |
| Team Invitations | Email invites | Access codes |
API Access
Both platforms provide full REST APIs.
API Capabilities
| Feature | Staminads | Umami |
|---|---|---|
| REST API | ||
| Feature Parity | Full access | Full access (all app operations) |
| Rate Limits | Configurable | 50 calls per 15 seconds |
| Authentication | API key | Bearer token or API key |
| Real-time API | Yes (30-minute window) |
Pricing
Staminads
Software cost (AGPL-3.0)
+ ClickHouse: Included or managed
- ✓ Unlimited pageviews
- ✓ Unlimited websites
- ✓ Unlimited team members
- ✓ Full API access
- ✓ All features included
Umami
Software cost (MIT License)
+ Database: $0 (Supabase free tier)
- ✓ Unlimited pageviews (self-hosted)
- ✓ Unlimited websites
- ✓ Full API access
- ✓ Basic funnel & journey reports
- ✓ Umami Cloud also available
Both Are Free to Self-Host
Both platforms are open source and free to self-host. Umami can run on Vercel + Supabase free tiers for $0/month. Staminads requires ClickHouse for optimal performance, which has some infrastructure cost. The choice comes down to feature differences: Staminads offers median metrics, advanced traffic categorization with backfill, and ad network detection. Umami offers basic funnels and journey reports (limited to same browser, same month).
Open Source
Open Source Comparison
| Feature | Staminads | Umami |
|---|---|---|
| Open Source | Yes (AGPL-3.0) | Yes (MIT License) |
| Source Code | Full access on GitHub | Full access on GitHub |
| GitHub Stars | Growing | 34,000+ |
| License Type | Copyleft (AGPL) | Permissive (MIT) |
| Technology Stack | Typescript + ClickHouse | Next.js + PostgreSQL/MySQL |
| Self-hosting | Docker | Docker, Vercel, various clouds |
License Difference
MIT (Umami): More permissive — you can use, modify, and distribute with minimal restrictions. Can be used in proprietary projects.
AGPL (Staminads): Copyleft — if you modify and distribute, you must share your changes. Network use counts as distribution. Ensures the project stays open source.
Unique Umami Features
Umami has some features that Staminads doesn't currently offer.
Features Unique to Umami
Reports
- • Basic funnel analysis (same browser/month)
- • Journey visualization (same browser/month)
- • Retention reports (same browser/month)
- • Attribution report
Features
- • Distinct IDs (cross-session tracking)
- • Run free on Vercel + Supabase
- • MIT license (more permissive)
- • 34K+ GitHub stars community
When Umami Might Be Better
If you need basic funnel analysis for same-browser, same-month journeys, Umami provides that today. If you want to run analytics for $0/month on free tiers, Umami works on Vercel + Supabase. If you need a permissive MIT license for proprietary use, Umami is the choice.
When to Choose Each Platform
Choose Staminads
Staminads is the better choice if you want to:
- Detect winning dimensions that drive engaged traffic
- Use median metrics (not misleading averages)
- Advanced traffic categorization to map your marketing mix
- Backfill historical data when rules change
- Auto-detect all ad networks (Google, Meta, TikTok, LinkedIn)
- No cross-session tracking (purest privacy stance)
Choose Umami
Umami might be better if you:
- Need basic funnel analysis (same browser/month only)
- Want basic journey/retention (same browser/month)
- Want to run for $0/month on free tiers
- Need MIT license for proprietary use
- Don't need advanced traffic categorization
- Don't need advanced data exploration
Frequently Asked Questions
Is Umami's fingerprinting really privacy-compliant?
It's legally contested. Umami hashes IP + User-Agent with rotating salts to identify visitors. They claim GDPR, CCPA, and PECR compliance. However, the EDPB Guidelines 2/2023 explicitly include fingerprinting within the scope of Article 5(3) of the ePrivacy Directive, which requires consent. Under GDPR, hashing is pseudonymization (not anonymization), and IP addresses are personal data. If you need to eliminate any legal gray area, Staminads' session-only approach avoids fingerprinting entirely.
Can I migrate from Umami to Staminads?
Yes. Both platforms track similar core metrics. Install the Staminads script and start collecting data immediately. Note that Staminads' different metric calculations (median vs average) and stricter privacy approach mean the numbers may not be directly comparable.
Why does Staminads use AGPL instead of MIT?
AGPL ensures that if anyone modifies and deploys Staminads, they must share their changes. This keeps the project open source and benefits the community. MIT is more permissive but allows proprietary forks. Both are valid open source licenses with different philosophies.
Why does Staminads require ClickHouse?
ClickHouse is a column-store database purpose-built for analytics — sub-second aggregation queries on billions of events. Umami uses PostgreSQL/MySQL (row-based databases), which are designed for transactional workloads and struggle with analytics at scale. As your dataset grows, Umami queries slow down significantly while Staminads stays fast. The tradeoff is slightly higher infrastructure cost.
Which has better community support?
Umami has a larger community with 34K+ GitHub stars, 300+ contributors, and an established ecosystem. Staminads is newer but growing. Both have active development and documentation.
Ready to find your winning dimensions?
Both Staminads and Umami are open source and self-hostable. Choose Staminads to detect which traffic sources, campaigns, and dimensions drive the most engaged visitors — with median metrics and advanced traffic categorization.