Understand why analytics numbers differ between platforms and how to reconcile discrepancies.
Why Numbers Differ
No two analytics tools show exactly the same numbers. This is normal and expected due to:
- Different methodologies
- Different tracking mechanisms
- Different definitions of metrics
- Different filtering and processing
Zenovay vs Google Analytics
Common Differences
| Metric | Zenovay | Google Analytics |
|---|---|---|
| Unique visitors | First-party cookie (cookie mode) or a server-side ID from a daily-salted hash (cookieless mode) | Client ID cookie |
| Sessions | 30-min inactivity (configurable, 5 to 120 min) | 30-min + midnight |
| Bounce rate | Visits with one page view | GA4: sessions that are not engaged (Universal Analytics: single-page sessions) |
| Page views | All loads | Sampled (free) |
Zenovay never uses browser fingerprinting. When cookieless mode is enabled, nothing is stored in the browser: the server derives an anonymous ID from a daily-salted hash, so the page views of one visit are linked, but a visitor who returns on another day counts as a new visitor. With cookies enabled, a first-party cookie recognises returning visitors across days instead. Either way the identifier is yours alone, and there is no cross-site fingerprint.
Why Zenovay Shows Fewer Visitors
1. Bot filtering:
- Zenovay filters known bots and automated traffic at ingest
- GA may count some bot traffic that Zenovay excludes
- Stricter bot filtering means fewer "visitors" in Zenovay
2. Privacy tools:
- Different script blocking rates
- Zenovay's lightweight script loads more often
- But some blockers target all analytics
3. Single-page apps:
- Virtual pageviews may be configured differently
- Check SPA tracking setup on both
Why Zenovay Shows More Visitors
1. Ad blockers:
- Zenovay's first-party tracking avoids many blockers
- GA blocked by default in some browsers
2. Cookie consent:
- GA requires cookies in strict mode
- Zenovay can track without cookies
3. Sampling:
- GA Free samples data at high volumes
- Zenovay counts every visit
4. Cookieless mode:
- The anonymous visitor ID changes every day
- A visitor who returns on another day counts as a new unique visitor
- Over longer periods this gives more unique visitors than a cookie-based tool
Reconciling Differences
Expect 10-30% variance as normal. To minimize:
-
Match date ranges exactly
- Include timezone considerations
- Use same start/end dates
-
Compare same metrics
- "Users" vs "Unique Visitors"
- "Sessions" definitions differ
-
Check for filters
- Both excluding internal traffic?
- Same geographic filters?
Session Count Differences
What's a Session?
Zenovay:
- Starts: First pageview
- Ends: after 30 minutes without activity (you can change this per site, from 5 to 120 minutes)
- Continues across subdomains with the same tracking code
- No midnight split in cookie mode. In cookieless mode the anonymous ID changes once a day, so a visit that is still going on at that moment can count as two
Google Analytics:
- Starts: First interaction
- Ends: 30 minutes inactive OR midnight OR campaign change
- Splits at midnight
Why Sessions Differ
Example scenario:
- User visits at 11:45 PM
- Stays until 12:15 AM
- Single continuous activity
Zenovay: 1 session (in cookie mode) GA: 2 sessions (split at midnight)
Aligning Sessions
If comparing:
- Compare full days, not partial
- Account for midnight splits
- Compare 7-day averages, not single days
Bounce Rate Differences
Definitions
Zenovay:
- Visits with exactly one page view ÷ all visits × 100
- Time on page, scrolling, clicks and events don't matter
- A visit with two page views is never a bounce, however short
Google Analytics 4:
- 100% minus the engagement rate
- A session is engaged when it lasts longer than 10 seconds, has a key event, or has 2 or more page views
- Only sessions that are not engaged count as bounces
Universal Analytics, Fathom, and Matomo or Plausible without custom events use the same single-page definition as Zenovay.
Why GA4 Usually Shows a Lower Bounce Rate
GA4 treats every session that lasts longer than 10 seconds as engaged, even if only one page was viewed, so it is not a bounce there. Zenovay only looks at page views, so the same one-page visit stays a bounce.
Example:
- A visitor lands on a blog post, reads for 2 minutes, leaves without opening another page
- Zenovay: Bounce (one page view)
- GA4: Not a bounce (engaged, longer than 10 seconds)
Neither number is wrong: they measure different things. Compare trends over time rather than the absolute values.
Pageview Differences
Counting Methods
Zenovay:
- Every page load counted
- Client-side + server-side deduplication
- Real-time processing
Google Analytics (Free):
- Sampled at high volume
- Session-based processing
- May have delayed counts
SPA Considerations
For single-page apps:
// Ensure both track virtual pageviews
// Zenovay
zenovay('page');
// GA
gtag('config', 'GA_ID', {'page_path': '/new-page'});
Geographic Data Differences
IP Geolocation
Different providers yield different results:
- MaxMind vs IP2Location vs custom
- Database update frequency
- VPN/proxy detection
Common issues:
- Country usually accurate
- City can vary 10-20%
- Mobile IPs often wrong location
Compare at Country Level
For valid comparison:
- Use country-level data
- Expect city data to vary
- Don't compare exact numbers
Real-Time vs Processed Data
Data Processing
Zenovay:
- Real-time display
- Immediate availability
- Live counts
Google Analytics:
- Processing delay (4-24 hours)
- Real-time view is separate
- Final numbers may change
When to Compare
Compare after both have finalized:
- Wait 24-48 hours after period ends
- Don't compare real-time to processed
- GA numbers may adjust over time
Conversion/Goal Differences
Goal Matching
Goals may differ due to:
- How goals count: a Zenovay goal counts visits, and each visit counts at most once per goal, even if the action happens several times. The conversion rate is visits that completed the goal divided by all visits in the selected period
- URL matching rules
- Event definitions
- Attribution windows
Check:
Zenovay: Is exactly "/thank-you"
GA: Contains "thank-you"
In Zenovay, upper and lower case, a trailing slash, www. and #anchors never matter in a goal's address, and the query string only matters if your pattern contains one.
Attribution Models
Conversion attribution:
| Model | Zenovay | GA |
|---|---|---|
| Default | Last-touch | Data-driven |
| Alternative | First-touch | First/last/linear, etc. |
Zenovay reports source attribution as last-touch by default, with first-touch available where the dashboard offers a model toggle. GA's data-driven model distributes credit differently, so conversion source counts won't line up one-to-one.
Traffic Source Differences
UTM Handling
Both use UTM parameters but may:
- Parse differently
- Store case differently
- Handle missing data differently
Standardize UTMs:
Use: utm_source=google
Not: utm_source=Google, GOOGLE, google.com
Referrer Detection
Differences in:
- Direct traffic definition
- Referrer header availability
- SSL/non-SSL referral
How to Investigate
Step-by-Step Comparison
-
Export both datasets
- Same date range
- Same timezone
- Raw data if possible
-
Compare totals first
- Calculate percentage difference
- 10-30% is normal
-
Drill into specifics
- Which days differ most?
- Which pages differ?
- Which sources differ?
-
Check configurations
- Both excluding same traffic?
- Both tracking same pages?
- Both using same events?
Using Debug Mode
Enable debug on both:
Zenovay:
<script data-debug="true" ...></script>
Check browser console:
- Both firing on same actions?
- Both receiving same data?
- Any errors on either?
Expected Variance
Normal Ranges
| Metric | Expected Variance |
|---|---|
| Pageviews | 5-15% |
| Unique visitors | 10-30% |
| Sessions | 15-25% |
| Bounce rate | 5-20% vs single-page definitions; GA4 is usually lower |
| Conversions | 10-20% |
When to Worry
Investigate if:
- Variance > 50%
- Sudden change in variance
- Trends go opposite directions
- One shows 0 data
Best Practices
Pick a Source of Truth
- Choose one platform as primary
- Use for business decisions
- Keep other for validation
Document Differences
Track known differences:
- Bot filtering rates
- Consent rate differences
- Technical setup differences
Focus on Trends
More reliable than absolute numbers:
- Week over week change
- Month over month growth
- Seasonal patterns
Both should show similar trends even if absolute numbers differ.
Contacting Support
If the variance still seems wrong, email support@zenovay.com with:
- Screenshots from both platforms
- The exact date range and timezone compared
- The specific metrics that differ
- Any recent changes to your tracking setup