Privacy vs. accuracy: what cookieless analytics can and cannot tell you

Privacy-focused analytics give you reliable, actionable data while respecting your visitors. But the way they work is fundamentally different from cookie-based tools. Understanding these differences helps you read your numbers correctly.

How cookieless identification works

Traditional analytics tools (Google Analytics, Adobe) place a persistent cookie in each visitor's browser. That cookie stays for months or years, making it easy to recognize returning visitors across sessions.

Cookieless analytics like Glyphex use a different approach. We generate a daily rotating hash based on anonymized browser properties: the visitor's IP address (never stored), user agent, and the current date. This means:

• The same person visiting today gets one identifier
• The same person visiting tomorrow gets a different identifier
• No personal data is stored or transmitted
• No consent banner is needed

This is the core privacy-accuracy tradeoff. You gain full compliance and complete data capture. In return, some metrics behave differently than you might expect from cookie-based tools.

What this means for your metrics

[Unique visitors](/glossary/unique-visitors): slightly overcounted

Because identifiers rotate daily, a person who visits your site on Monday and again on Tuesday is counted as two unique visitors. Cookie-based tools would count them as one.

In practice: For most sites, daily unique visitor counts are accurate. Weekly and monthly totals will be higher than what cookie-based tools would report, typically by 10-30% depending on how often your visitors return.

Why it still works: The trend is what matters. If your visitors go up 20% week over week, that 20% change is real regardless of the absolute number. Relative comparisons, trends, and proportions remain accurate.

Sessions: accurate within a day

Sessions are tracked using a session window (typically 30 minutes of inactivity). Within a single day, sessions are accurate. A visitor who browses your site, leaves for lunch, and comes back is correctly counted as two sessions.

Across days, session continuity is lost. A visitor who starts browsing at 11:55 PM and continues past midnight would start a new session. This is a rare edge case that doesn't meaningfully affect data quality.

Pageviews: fully accurate

Every page load is counted exactly once. There is no privacy tradeoff here. Your pageview numbers are precise.

[Bounce rate](/glossary/bounce-rate): accurate

Bounce rate is calculated per session. Since sessions within a day are accurately tracked, bounce rate is reliable. A visitor who views one page and leaves is correctly counted as a bounce.

Traffic sources: fully accurate

Referrer information, UTM parameters, and channel classification don't depend on visitor identification. These metrics are just as accurate in cookieless analytics as in cookie-based tools.

Geographic and device data: fully accurate

Country, city, browser, device type, operating system, and screen size are captured from each request. No identification is needed, so these metrics are precise.

Returning visitors: not available

Cookie-based tools can tell you how many visitors are "new" vs. "returning" over a long period. With daily rotating identifiers, this distinction isn't possible beyond a single day. This is an intentional tradeoff: tracking returning visitors across days requires persistent identifiers, which means cookies and consent banners.

How to read your dashboard

Keep these principles in mind:

Focus on trends, not absolutes

A metric going up or down by a certain percentage is meaningful and accurate. The exact absolute number may differ from what a cookie-based tool would show, but the direction and magnitude of change is reliable.

Compare within the same system

Don't compare Glyphex numbers directly with Google Analytics numbers. They measure differently. Compare Glyphex data with Glyphex data over time.

Shorter time ranges are more accurate

A "Last 24 hours" view is highly accurate because identifiers don't rotate within a day. As you extend to 7, 30, or 90 day ranges, unique visitor counts become increasingly overcounted relative to cookie-based tools.

Pageviews and sessions are your most reliable metrics

These don't depend on cross-day identification and are accurate regardless of time range.

The bigger picture: what you actually lose vs. gain

What you lose

• Precise multi-day unique visitor deduplication
• New vs. returning visitor segmentation (beyond a single day)
• Individual user journey tracking across days

What you gain

• 100% data capture (no visitors lost to consent declines)
• No consent banners interrupting user experience
• Full GDPR, CCPA, and ePrivacy compliance without legal complexity
• Accurate data from privacy-browser users (Safari, Firefox, Brave) who block cookies
• Sub-1KB tracking script that doesn't slow your site

For most websites, the data you gain from capturing every visitor far outweighs the precision you lose from rotating identifiers. Cookie-based tools often miss 30-50% of visitors who decline consent or use ad blockers. That's a much larger source of inaccuracy than daily identifier rotation.

Best practices

1. Use pageviews or sessions for absolute comparisons when you need exact numbers (e.g., for reporting to clients or stakeholders)
2. Use visitor trends for growth analysis — the percentage change is accurate even if the absolute count is slightly inflated
3. Use shorter time ranges when precision matters most
4. Don't mix data sources — pick one analytics tool and use it consistently
5. Remember that all analytics are estimates — even cookie-based tools face challenges with cross-device tracking, browser restrictions, and ad blockers