Identify Visitors Without Cookies
Track complete user journeys with typically 95-99.5% accuracy in matching returning visitors on the same device/browser—no cookies needed. Based on testing across millions of visitor sessions.
Accuracy Definition: 99.5% precision (avoiding false matches) with 98%+ recall (catching return visitors). Accuracy varies by browser diversity—typically 95-99.5% depending on your audience's device/browser combinations.
Accuracy Breakdown by Scenario
Real-world accuracy varies based on device and browser conditions
Same Device, Same Browser
Optimal conditions—highest accuracy
Same Device, Browser Updated
ML models adapt to minor changes
Same Device, Privacy Mode
Reduced accuracy in incognito sessions
Shared Devices
Family computers, kiosks, public devices
Comparison Metrics
- • 3-5x more accurate than IP-based identification
- • Comparable to first-party cookie tracking without the consent complexity
- • Collision rate under 0.1% for sites with diverse traffic
Fingerprint Stability
- • ML models identify return visitors even when 20-30% of fingerprint characteristics change
- • Behavioral patterns help confirm identity when fingerprint partially changes
- • Fingerprint matching is robust to minor browser/OS updates
Technical Transparency
Understanding what browser characteristics we use and how fingerprinting works
Browser Characteristics Used
We create fingerprints using only basic, publicly available browser characteristics—no invasive tracking methods:
- • Browser type and version (Chrome, Firefox, Safari, etc.)
- • Screen resolution (standard web API)
- • Timezone and language settings
- • Page views and navigation paths
- • Click and scroll interactions
- • Session duration and timing
What we don't use: Canvas fingerprinting, audio fingerprinting, WebGL fingerprinting, installed fonts detection, or hardware characteristics. We explicitly avoid techniques that privacy advocates consider invasive.
Fingerprint Example
Visitor ID
a8f3b2c1-d4e5-...
First Visit
Jan 15, 2024, 2:30 PM
Last Visit
Feb 7, 2024, 10:15 AM
Total Visits
5
Browser
Chrome
Country
United States
Theme
light
Language
en
Confidence this visitor matches a previous visit
How Fingerprinting Works
1. Fingerprint Creation
On first visit, we collect browser characteristics and create a unique fingerprint hash. This hash is one-way encrypted and cannot be reverse-engineered to identify the user.
2. Fingerprint Matching
On return visits, we compare the new fingerprint to stored fingerprints. ML models use fuzzy matching to identify return visitors even when 20-30% of characteristics change (e.g., browser updates, OS updates).
3. Cross-Device Limitations
Important: Fingerprinting works per device/browser combination. We cannot track across completely different devices without login events. If a user visits from a laptop and then a phone, they appear as two separate visitors unless they log in.
Privacy & Compliance
Transparent about what we collect and how it's used
What We Don't Collect
We create pseudonymous identifiers—not personal data. Specifically:
- • No names, emails, or directly identifying information
- • No IP addresses stored or used for identification
- • No cookies or local storage
- • No cross-site tracking—fingerprints are site-specific
GDPR & Privacy Compliance
Important Legal Note: GDPR applicability depends on how data is used, not just collection method. While WysLeap is designed to minimize personal data collection:
- • Fingerprinting for analytics purposes may still require disclosure in your privacy policy
- • GDPR compliance depends on your specific use case and jurisdiction
- • We recommend consulting legal counsel for your specific situation
- • Some regulators consider fingerprinting similar to cookies and requiring consent under ePrivacy Directive
Our Approach: We recommend a privacy-first approach with transparency. Even if consent isn't strictly required, disclosing fingerprinting in your privacy policy builds trust with users.
Privacy Mode Clarification
When we say "works in privacy mode," we mean:
- • Fingerprinting works across normal sessions despite privacy settings
- • Within a single incognito session, fingerprinting works normally
- • Across incognito sessions, accuracy is reduced (85% vs 99.5%) because incognito mode clears some fingerprinting data between sessions
- • Most fingerprinting has reduced accuracy in private browsing—we're transparent about this limitation
Data Retention
We retain visitor identifiers and fingerprints for 24 months of inactivity. After 24 months without visits, visitor data is automatically purged. Active visitors' data is retained as long as they continue visiting.
Privacy-Preserving Machine Learning
How ML improves accuracy while protecting privacy
What the ML Actually Does
- Stability Learning: Models learn which fingerprint characteristics are most stable over time (e.g., screen resolution rarely changes, but browser version does). This helps identify return visitors even when some characteristics change.
- Fuzzy Matching: Algorithms identify return visitors despite minor fingerprint changes (browser updates, OS updates). ML models can match fingerprints that are 70-80% similar, not just exact matches.
- Behavioral Pattern Analysis: When fingerprint partially changes, behavioral patterns (visit frequency, page preferences, navigation patterns) help confirm identity.
What Makes It Privacy-Preserving
- Aggregated Pattern Learning: Models train on aggregated patterns across all visitors, not individual visitor data. No raw fingerprints leave your infrastructure.
- One-Way Hashing: Identifiers are one-way hashed and cannot be reverse-engineered to identify the user or reconstruct the original fingerprint.
- Site-Specific: Fingerprints are unique to your site. We don't share fingerprints across sites or create cross-site profiles.
- No Personal Data: ML models never see or learn from personal data—only anonymous browser characteristics and behavioral patterns.
Real-World Use Cases
Specific scenarios where visitor identification drives value
Scenario 1: Multi-Session Research Behavior
B2B buyers typically visit 5-7 times before converting. See their complete research journey to understand what content drives decisions.
Scenario 2: Cart Abandonment Follow-Up
Identify return visitors who abandoned carts to understand if they're comparison shopping or lost interest. Track whether they return to complete purchases or browse different products.
Scenario 3: Content Attribution
Which blog posts lead to conversions 3 weeks later? Track the complete path from content discovery to purchase, even when visitors don't convert on their first visit.
Cookies vs. Fingerprinting
| Metric | Cookies | Fingerprinting |
|---|---|---|
| Accuracy (same device) | 98-99% | 95-99.5% |
| Works in private browsing | No | Limited (85%) |
| Requires consent (EU) | Yes | Depends* |
| User can block | Easily | Difficult |
| Cross-browser tracking | No | No |
| Data persistence | User-controlled | Server-side |
* GDPR/ePrivacy interpretation varies by jurisdiction. Consult legal counsel for your situation.
Known Limitations
Building trust through transparency about what we can and cannot do
Shared Devices
Less accurate on shared devices (family computers, kiosks, public devices). Multiple users on the same device may appear as a single visitor.
Private Browsing
Reduced accuracy (85% vs 99.5%) in private browsing modes. Incognito sessions clear fingerprinting data between sessions.
Cross-Device Tracking
Cannot track across completely different devices without login. A user visiting from laptop and phone appears as two separate visitors.
VPNs & Network Changes
Users with VPNs or frequently changing network configurations may appear as new visitors if fingerprint characteristics change significantly.
Mobile Apps
Mobile app tracking requires different approach. Fingerprinting works for mobile web browsers, but native apps need device IDs or other methods.
Anti-Fingerprinting Measures
Some browsers actively resist fingerprinting (Firefox Enhanced Tracking Protection, Safari ITP, Brave randomization). We adapt to these measures, but accuracy may be reduced.
How WysLeap Compares
vs. Cookie-Based Tracking
More resilient to cookie deletion, works in stricter privacy contexts. Comparable accuracy (95-99.5% vs 98-99%) without requiring consent banners in many jurisdictions.
vs. IP-Based Tracking
Much more accurate (3-5x improvement). Handles dynamic IPs and shared networks better. IP addresses change frequently and are shared by multiple users—fingerprinting provides device-level identification.
vs. Login-Required Tracking
Works for anonymous visitors, no authentication required. Enables multi-session analytics for visitors who haven't signed up yet—critical for understanding conversion funnels.
vs. No Tracking
Enables multi-session analytics without compromising privacy excessively. Provides insights into user journeys that single-session tracking cannot capture.
How Does This Compare to Google Analytics?
GA4 uses cookies + consent-based tracking: Google Analytics 4 relies on cookies and requires consent banners in the EU. When users reject cookies or use private browsing, GA4 loses tracking.
WysLeap uses fingerprinting: Different trade-offs—more resilient to cookie blocking but may have different privacy implications. Fingerprinting works even when cookies are blocked.
Can be used together: Many customers use GA4 for basic metrics (with consent) and WysLeap for cookieless journey tracking. They complement each other—GA4 provides Google's ecosystem integration, WysLeap provides privacy-resilient visitor identification.
Responsible Use
WysLeap's Visitor Identification is Designed For:
- • Analytics and understanding user behavior
- • Improving user experience
- • Multi-session journey tracking
- • Conversion funnel analysis
We Explicitly Prohibit Use For:
- • Cross-site tracking without disclosure
- • Selling or sharing fingerprint data with third parties
- • Using fingerprints to personally identify individuals
- • Any use that violates user privacy expectations
Privacy Policy Disclosure
We recommend disclosing fingerprinting in your privacy policy, even though it doesn't collect traditional personal data. Transparency builds trust with users and helps ensure compliance with privacy regulations.
For technical details on our fingerprinting methodology, see our technical documentation or contact support.
Frequently Asked Questions
Is fingerprinting legal?
It's a gray area that varies by jurisdiction. In the EU, fingerprinting may be considered similar to cookies under ePrivacy Directive. In the US, it's generally legal but subject to state privacy laws. We recommend consulting legal counsel for your specific jurisdiction and use case.
Can users opt out?
Yes. Users can opt out through browser settings (Firefox Enhanced Tracking Protection, Safari ITP) or by using privacy-focused browsers. We respect browser-level privacy controls and don't attempt to circumvent them. Some customers also provide opt-out mechanisms in their privacy policies.
Can users delete their visitor profile?
Yes. Users can delete their visitor profile and all associated data at any time. When a profile is deleted, it permanently removes the visitor profile data, all visit history, and all associated telemetry events. Users can access this feature on our privacy page where they can view their visitor profile and delete it if desired.
How is this different from third-party tracking?
WysLeap fingerprinting is first-party—fingerprints are created and stored only for your site. We don't share fingerprints across sites or create cross-site profiles. This is fundamentally different from third-party tracking networks that track users across multiple websites.
What happens on shared devices?
Accuracy is reduced (around 70%) on shared devices. Multiple users on the same device may appear as a single visitor. This is a known limitation of fingerprinting—it identifies devices/browsers, not individual users.
Does this work on mobile?
Yes, for mobile web browsers. Fingerprinting works on mobile Safari, Chrome Mobile, and other mobile browsers. However, native mobile apps require different approaches (device IDs, advertising IDs) as fingerprinting is primarily a web browser technology.
How do browser updates affect tracking?
ML models are designed to handle fingerprint evolution. When browsers update, 20-30% of fingerprint characteristics may change, but our fuzzy matching algorithms identify return visitors despite these changes. Behavioral patterns also help confirm identity when fingerprints partially change.
Start Tracking Complete User Journeys
Get typically 95-99.5% identification accuracy without cookies. See live examples of fingerprint characteristics, multiple sessions being connected, and journey visualizations.