You added a cookie banner, and your GA4 numbers dropped the same week. It looks like readers left. The more likely cause is that GA4 can no longer see the people who decline the banner or ignore it.
Those visitors still load your pages and read your posts. GA4 just doesn’t record them. So before you rewrite headlines or blame a Google update, find out how much of the drop is missing data. You can measure that on your own site in a couple of weeks.
What GA4 sees behind a banner
How much GA4 loses depends on how Consent Mode is set up. Google’s Consent Mode help page describes two ways to run it.
In the basic setup, Google tags are blocked until the visitor consents. A visitor who clicks decline, or never touches the banner, sends GA4 nothing, so GA4 has no record of the visit. Google says that with basic you “will not get modeled data” to fill the gaps.
In the advanced setup, Google tags load before the consent dialog appears. When a visitor denies consent, the tags send cookieless pings to Google instead of storing cookies. Google uses those pings for behavioral and conversion modeling, which fills in part of what’s missing.
Either way, the people who say no or never answer are the ones who go missing. Which setup your banner plugin uses decides whether Google tries to estimate them.
What Consent Mode gets back, and what it doesn’t
Advanced mode sounds like a full fix. Google’s page on behavioral modeling shows it comes with conditions.
Consent mode has to run on all pages, and Google tags must load in all cases, not only when the visitor consents. Your property also has to meet two thresholds:
- At least 1,000 events per day from visitors who denied analytics consent (analytics_storage denied), for at least 7 days.
- At least 1,000 daily users who granted analytics consent, on at least 7 of the previous 28 days.
Meeting those thresholds doesn’t guarantee eligibility. You must also pick “Blended” under Reporting identity to see modeled data.
Modeling also covers only some numbers: users, sessions and new users. It isn’t applied to event counts like page_view. It isn’t supported in audiences, retention reports, user explorer, cohort explorations, or data export, including BigQuery.
A site under those thresholds gets no modeled data at all, and the visitors who declined are absent from its reports.
How a cookieless tool counts the people GA4 misses
To see what the banner hides, you need a second count that doesn’t stop at the banner. Vireo, the free analytics plugin from RebelCode, counts visits without waiting for consent by default. Its privacy documentation sets out what it stores.
Vireo sets no cookies and doesn’t use local storage for tracking, so nothing is stored on visitors’ devices.
To tell visitors apart within a day, it combines the IP address and browser with a secret that changes every day, and keeps only a scrambled result. It holds today’s and yesterday’s secrets and deletes older ones. Once a day’s secret is gone, nobody can work out which address a result came from, and a visitor can’t be followed from one day to the next.
Raw IP addresses aren’t stored in the analytics tables. The IP sits in a short-lived file while a view is processed, and Vireo uses it to look up the country. That file is deleted when the view reaches the database, usually within a minute. All data stays in your own WordPress database. Visitor data isn’t sent to RebelCode or anyone else.
Vireo also skips some people on purpose. Visitors whose browser sends Do Not Track or Global Privacy Control aren’t counted at all. There’s no setting to change that. Logged-in administrators aren’t counted by default, and bots are kept out of the totals and reported separately.
Pageviews are exact over any range. Visitors are a true unique count within one day. Over longer ranges, Vireo adds up each day’s unique visitors. The what-counts page has the details. For the wider picture, see cookieless WordPress analytics.
What the rules say
Whether a tool needs consent is a legal question, and the answer varies by country.
Article 5(3) of the ePrivacy Directive allows storing information on a user’s device, or gaining access to information already stored there, only with consent. The exception is where it’s strictly necessary to provide a service the user asked for, or for transmitting a communication.
The EDPB, the EU’s board of data protection regulators, adopted Guidelines 2/2023 on the technical scope of Article 5(3) in October 2024. They say the rule isn’t limited to cookies. JavaScript that tells a browser to send information back to a server can count as “gaining access.” Tracking based on the IP address can fall under it too. The same guidelines say that applying Article 5(3) doesn’t automatically mean consent is needed. Each case has to be assessed against the exemptions.
Some national regulators have written those exemptions down for analytics. France’s CNIL exempts audience measurement from consent when it’s strictly limited to measuring the site’s audience for the publisher alone. It must produce only anonymous statistics. It can’t be cross-referenced with other data or passed to third parties, and it can’t track a person’s browsing across sites.
In the UK, the ICO’s guidance has a statistical purposes exception. The sole purpose must be collecting aggregate statistics to improve the service. Users get clear information about it and a simple, free way to object. A third-party analytics provider must act only for you, as a processor.
Rules differ by country, and you decide with your own adviser whether your setup needs consent. Vireo’s docs say they aren’t legal advice, because the notice you need depends on your site and where you are. They recommend mentioning Vireo in your privacy policy.
If you run a consent plugin, Vireo works with those that use the WP Consent API, such as Complianz, Cookiebot and Real Cookie Banner. It registers with the WP Consent API as a plugin that respects consent. By default it counts visits without waiting for consent, because it stores nothing on the visitor’s device. A site that wants Vireo to wait for statistics consent adds one filter:
add_filter( 'vireo_require_consent', '__return_true' );
The filter only takes effect while a consent plugin is active. The check runs in the visitor’s browser, so it works with page caching. The consent plugins section of the docs covers the setup.
Measure your own gap
How much a banner hides depends on its design, your audience and your country. Only a measurement on your own site gives you your number.
- Install Vireo and leave GA4 and the banner exactly as they are. Leave
vireo_require_consentoff while you measure, if your adviser is comfortable with that. Turned on, it stops Vireo counting people who decline, and you lose the comparison. - Let both tools run for a couple of weeks.
- Compare pageviews, one day at a time. Take GA4’s page_view event count from its standard reports and set it against Vireo’s pageviews for the same day. Google doesn’t model page_view counts, so the GA4 figure is what GA4 actually observed. Skip visitors and users: the two tools define them differently, as this post on why analytics tools show different numbers explains. Set both tools to the same timezone, or compare multi-day totals to smooth out the boundary.
- Divide GA4 pageviews by Vireo pageviews. The result is roughly the share of views GA4 sees. One minus that share is roughly what the banner, plus ad blockers, hides.
Say GA4 shows 600 pageviews on a day Vireo shows 1,000. That’s 0.6, so GA4 sees about 60% of views and about 40% are hidden. Those figures are made up to show the arithmetic, and yours will differ.
Reading the number
Treat the result as an estimate, because several things push it in different directions.
Blocking lists used by many ad blockers, such as EasyPrivacy, block Google Analytics’ domains. That widens the gap for reasons that have nothing to do with your banner. Vireo leaves out Do Not Track and Global Privacy Control visitors, admins and bots, which narrows it. GA4 counts you unless you filtered yourself out. Watch whether the ratio holds steady from day to day. A steady ratio is more useful than any single day’s figure.
Once you have your number, you have two reasonable paths. You can run both, and read traffic trends from the tool that counts everyone. Or you can decide GA4 isn’t worth the banner.
If GA4 was the only reason you added the banner, check with your adviser whether you still need one. Other plugins, ads and embeds may set cookies too.