If you have ever opened Google Analytics 4 next to another analytics tool and seen two different numbers, you are not doing anything wrong. The two tools count differently by design. GA4 leans on cookies and on consent: when a visitor declines the cookie banner, GA4 stops collecting and fills the gap with modelled, estimated data. It also applies thresholding and sampling, so small segments get rounded or withheld. A cookie-free tool like Ghaze measures every visit without a cookie or a consent banner, so there is nothing to model around and nothing to suppress. That is the headline reason the two disagree, and it shows up most in conversion and goal counts, where a handful of missed events changes the whole picture. This post walks through where GA4 and Ghaze diverge, why, and how to read each number so you can trust the one you act on.
Consent gaps and modelled data
GA4 runs on Google's consent mode. When a visitor declines the cookie banner, GA4 stops collecting measured hits for that visit and fills the gap with modelled, estimated data instead. The model is a reasonable guess, not a record of what actually happened.
Ghaze is cookie-free and needs no consent banner, so there is no gap to model around: every visit is measured the same way. That single difference explains most of the disagreement you will see between the two tools, and it is largest on sites where a big share of visitors decline cookies.
Cookie loss, ITP and Private Relay
Even when a visitor accepts cookies, those cookies do not always last. Safari's Intelligent Tracking Prevention, Firefox's Enhanced Tracking Protection and Apple's iCloud Private Relay shorten or strip the identifiers that GA4 and pixel-based conversion tracking depend on.
Coverage degrades quietly over time, so a returning visitor can be counted as new, or missed altogether. This is a well-documented, industry-wide trend rather than anything specific to one site. A measured, cookie-free approach does not lean on those identifiers in the first place.
Thresholding and sampling
To protect individual identity, GA4 withholds or rounds low-volume rows, a behaviour known as data thresholding, and it samples large date ranges rather than counting every row. Both are sensible privacy and performance choices.
The catch is that the number you read is not always the number that happened, especially for small segments and long reporting windows. Ghaze reports rows as measured, so a small but real segment does not disappear behind a threshold.
Goals and phone-call counting
Goals like form submissions, email clicks and phone calls are where small collection gaps turn into a visible difference. Each one is a low-count, high-value event, so losing even a few changes the picture in a way that high-traffic pageviews never would.
If you bid or bill on those events, the gap matters: it flows straight into your cost-per-lead and into the signal your Google Ads bidding learns from. This is exactly the category where a consent gap or a stripped cookie does the most damage to GA4's count.
How to read each number
A practical rule: use GA4 for broad trends and channel-level direction, and use a measured, cookie-free count for the conversions you bill or bid on. When the two disagree on a goal that matters, trust the tool that measured every visit over the one that modelled around a missing banner.
Ghaze is built for that job: one lightweight script in place of GA4, Hotjar and a separate uptime monitor, privacy-friendly, with your data hosted in Sydney. Switching does not mean starting from zero either, because you can bring your GA4 history across on the way in.
/ where to go next
- the full GA4 vs Ghaze comparison: The head-to-head breakdown, including the honest cases where GA4 is the better pick.
- GA4 import: Bring your GA4 sessions, conversions and revenue across so you keep your history.
- conversion tracking: Forms, calls, emails and sales detected automatically, with no tags or Tag Manager.
- goal tracking: Phone-call and form-submission goals counted as they happen.
- security and data residency: Privacy-first by default, with your data hosted in Sydney, Australia.
- real-user web vitals vs Lighthouse: Why your speed score and Search Console disagree, and the 75th-percentile field data Google actually ranks on.
Frequently asked questions
Why do GA4 and Ghaze show different numbers?
Mostly because GA4 depends on cookies and consent. When a visitor declines the cookie banner, GA4 models the missing data and can also threshold or sample low-volume rows. Ghaze is cookie-free and needs no banner, so it measures every visit directly. The two methods produce different counts, and the gap is widest on conversions and goals.
Which number should I trust for conversions?
For the conversions you bill or bid on, trust the measured, cookie-free count. GA4 is useful for broad trends and channel direction, but a single modelled or thresholded gap can move a low-count, high-value event like a phone call or form submission enough to mislead your reporting and your ad bidding.
Can I bring my GA4 history into Ghaze?
Yes. Ghaze offers a GA4 import so you can migrate your existing sessions, conversions and revenue without losing your history when you switch.
Is Ghaze cookie-free?
Yes. Ghaze is cookie-free and privacy-friendly, so it does not require a cookie consent banner, and your data is hosted in Sydney, Australia.