Tutorial · 25 August 2026
Leave Google Analytics on. Add a 1 KB script. Compare totals.
GA4 is built for ads identity. Cookie Free Analytics is a 1 KB script and one dashboard. The useful week is not a cutover. It is both tags on the same pages.
Run both and compare your own two totals for the same week. The difference is mostly people who clicked Reject, plus lists that drop googletagmanager.com. That is the number worth a meeting — not a migration project.
What not to compare
Do not compare GA “active users” to our visitors. Ours collapse nine refreshes with a daily HMAC and forget you tomorrow. GA is a graph for ads. Pageviews are the overlapping fact. Sources are next, if your UTMs are the same five: source, medium, campaign, content, term.
Keep GA if you sell ads on the back of the visit. Add this if you wanted to know whether /pricing moved after a tweet.
The steps
- 1
Leave GA where it is
Do not turn consent mode off as an experiment. You are measuring the hole, not cleaning up GA.
- 2
Add Cookie Free Analytics
Create a site, paste the 1 KB snippet. First-party is a CNAME on a subdomain such as c.yourdomain.com if the hostname is blocked. Other hosts: default snippet first, first-party in week two.
- 3
Wait seven UTC days
Compare pageviews, not “users.” GA users are an ads identity. Ours are distinct HMAC tokens per local day. The comparable number is pageviews, then sources.
- 4
Read the gap
If our total is higher, the usual causes are Reject, list blockers, and Safari. If GA is higher, check our hostname filter, a missing snippet on a template, or a SPA you have not hooked — we count pushState.
Questions
- Will you import my GA history?
- No. Both run side by side from today. We do not pretend a property ID is a pageview archive we can swallow.
- Should I delete GA after a week?
- Only if you do not need ads audiences, remarketing, or BigQuery joins. If you need those, keep GA. This script is pageviews, sources, and goals.