EMPIREYO

← Knowledge base

Online marketing

Cookieless measurement: stop guessing on half the data

Your campaigns run and your dashboard shows numbers, but part of your visitors is missing. Anyone who refuses tracking cookies should not be in your measurement. This article explains the rules, what techniques such as Consent Mode and server-side tagging do and do not fix, and how to measure without tracking.

 ·  by  ·  Reading time 13 min  ·  Online marketing  ·  updated 12 November 2026

In short

Cookieless measurement means measuring without tracking cookies, or with as few as possible. In the Netherlands you may only place tracking cookies after consent, and refusing must be as easy as accepting. Anyone who refuses is missing from your data. You do not fix that gap with a trick. Choose a privacy-friendly analytics tool, measure after consent, or count what actually comes in.

Two chains, one budget

Same ad, same click. The difference is what happens to the measurement afterwards.

Leaking chainClick on ad→Browser blocks measurement→Conversion invisible→Algorithm guesses
Working chainClick on ad→Server-side tag measures→Conversion returns in full→Algorithm learns on real data

The drawing above shows two chains: a measurement that leaks along the way and one that comes back complete. That is the technology. But which route you may choose depends first on something else: what the visitor has allowed. A measurement that catches everything because it ignores the visitor's choice is not a closed chain but an infringement.

Cookieless is used in three ways. Sometimes it means no cookies at all. Sometimes it means only your own cookies from your own domain, none from other companies. And at Google it also means measurements sent without cookies when someone does not consent. Legally, those three are not the same.

In this article you will read what the law says, what the Dutch Data Protection Authority asks of a cookie banner, why your measurement misses part of your visitors, which routes there are, what Consent Mode and server-side tagging do and do not do, how to use your own data and how to measure what counts.

What does the law say about cookies and measurement?

In the Netherlands the rule is in Article 11.7a of the Telecommunications Act. You may only store information on a visitor's device, or read information from it, if you have informed them clearly and fully and they have given consent. There are exceptions: when it is only needed to make the communication possible, when it is strictly necessary for the service the visitor asks for, or when it concerns information about the quality or effectiveness of your site with no or minor consequences for privacy.

The Dutch Data Protection Authority (AP) turns that into three kinds of cookies:

KindWhat forConsent needed?
Functional cookiesMaking the site work: logging in, language choice, basketNo, but you must say that you place them
Limited analytics cookiesVisitor numbers and the most visited pagesNo, if the consequences for privacy are none or minor
Other cookies, such as tracking cookiesFollowing behaviour, also across other sites, and building profilesYes, in advance

Two points are easily missed. The rule does not only apply to cookies. According to the AP, the same rules of thumb apply to every technique that stores something on or reads something from the device, such as local storage, tracking pixels, web beacons and fingerprinting. And if limited analytics cookies process personal data, you still need a legal basis under the GDPR, and it belongs in your privacy statement.

This is general information, not legal advice. If you are unsure about your own situation, ask a lawyer.

What does the AP ask of a cookie banner?

The AP gives nine rules of thumb for a clear cookie banner. In short:

  1. Say which personal data you process and why, before anyone chooses.
  2. Do not tick boxes in advance, not even in a second layer.
  3. Use clear words on the buttons, such as accept and reject.
  4. Put reject and accept on the same layer.
  5. Do not hide the reject button, for example by letting it blend into the background.
  6. Do not make rejecting take more clicks than accepting.
  7. Do not tuck rejecting away as an inconspicuous link in the text.
  8. Make withdrawing consent as easy as giving it.
  9. Do not confuse consent with legitimate interest.

A cookie wall, where you cannot use the site normally if you refuse, is prohibited according to the AP. Carrying on scrolling does not count as consent.

According to the AP, the first layer of the banner should at least say that you process personal data, who does so and why, whether you share data with other parties and with how many, and that the visitor can withdraw consent. If you use an off-the-shelf banner tool, take care: the AP writes that their default settings often do not comply with the GDPR in practice. The responsibility stays with you, not with the supplier.

The AP has checked more strictly since 2024. In recent months it warned around a hundred organisations about their cookie banner, including in media, healthcare and insurance. Most of them then changed the banner. On 16 October 2026 the AP announced an investigation into two organisations that, after explanation and a warning, still had no directly visible reject button. The first cookie fine went earlier to the company behind the Dutch chemist chain Kruidvat, for tracking cookies without valid consent.

Why does your measurement miss part of your visitors?

Because that is how it should be. You may not track someone who refuses a tracking cookie. An analytics tool that asks for consent therefore does not see those visitors, or only as loose, anonymous counts. On top of that, some browsers and browser extensions block measurement scripts, regardless of what someone chooses in the banner.

How big the gap is differs by site, by audience and by banner. A figure from someone else's study says little about your site. So measure it yourself: put the number of enquiries your analytics tool shows next to the number of enquiries that actually arrived in your inbox or records, over the same month. The difference is your gap. That figure is more useful than any estimate.

The gap is not evenly spread either. People who refuse may behave differently from people who accept. A channel can look weaker than it is as a result. So use your measurement mainly for trends and comparisons within your own site, and less as an exact count.

Which routes are there?

There is no single right answer. There are routes, each with its own price:

RouteWhat it isConsentWhat you miss
Not measuring on the siteOnly counting what comes in: enquiries, calls, salesNot neededHow visitors behave on the site
Privacy-friendly analytics toolStatistics without following visitors across the internetNot needed if the consequences for privacy are none or minorProfiles and links with advertising
Google Analytics after consentThe script only loads after a yesIn advance, through the bannerEveryone who refuses
Advanced Consent ModeThe script always loads and sends measurements without cookies when there is no consentFor cookies in advance; see belowLess than the previous route, but only as an estimate
Server-side taggingMeasurement runs through a server you manageThe same rules as aboveDepends on what you pass on

As an alternative to intrusive analytics, the AP itself mentions services that keep statistics without following visitors across the whole internet, giving Matomo, Apache Superset, Plausible Analytics and OpenPanel as examples, provided they are set up properly. For advertising, the AP also points to an alternative: contextual ads, which look at the content of the page rather than at the person.

And the AP makes a point that is often forgotten: many organisations collect data they never use. If you stop collecting personal data from visitors, you no longer need to ask for consent and you can remove the cookie banner.

What does Google's Consent Mode do?

Consent Mode is how Google tags take the visitor's choice into account. Google distinguishes two forms:

  • Basic. The Google tags only load once the visitor makes a choice in the banner. If they say no, nothing goes to Google, not even the choice itself.
  • Advanced. The tags load straight away. As long as there is no consent, Google says they send measurements without cookies, including a timestamp, the browser and the previous page. After a yes, the full measurement follows.

With advanced, Google Analytics can estimate the behaviour of people who refused using a model. But there are thresholds. According to Google, your site then needs at least a thousand events a day without consent for at least seven days, and at least a thousand users a day with consent on seven of the past 28 days. Even then the model is not certain to work. If your site does not reach those numbers, you will not see estimated figures with advanced either.

Whether measurements without cookies may be sent before consent is a legal question we cannot answer for you. The AP writes that its rules of thumb apply to every technique that stores information on or reads it from the device, including tracking pixels. So put the choice for advanced to a lawyer before you make it.

Does server-side tagging solve it?

With server-side tagging, such as the server container of Google Tag Manager, the browser first sends the measurement to a server you manage. According to Google, you have full control there over how the data is shaped and where it goes. So you can remove fields before they go to another company, and you decide which services receive anything.

That is a real advantage for privacy and control. But it changes nothing about the rule: if you set a cookie or read the device for tracking through that route, you still need consent. Server-side tagging is a way to measure properly after consent, not a way around consent. And it adds a server that someone has to set up, maintain and secure.

What is first-party data, and what may you do with it?

First-party data is data you collect yourself: enquiries, quotations, customers, newsletter sign-ups. It is valuable because it is about real customers and does not depend on a cookie. But it is still personal data, with all the rules of the GDPR.

One example is enhanced conversions in Google Ads. You send data a customer entered in a form, such as an email address, scrambled with a hash function, to Google. Google matches it to Google accounts that were signed in when someone engaged with your ad. According to Google, this only works if the visitor has consented to sharing user data for advertising. So say in your banner and your privacy statement that you do this, and with whom.

The most useful way to use your own data is often the simplest: counting. How many enquiries came in, from which region, for which service, and how many became customers? That does not need a single cookie.

How do you measure what counts, without tracking?

  1. Count enquiries at the source. Every enquiry arrives in an inbox or system. Record how many there were each month.
  2. Ask how people found you. A choice field in your form gives an honest picture, including word of mouth that no analytics tool sees.
  3. Use Search Console for what Google knows about your search traffic. Those figures come from Google itself, not from a script on your site.
  4. Measure campaigns with their own addresses, such as a separate landing page per campaign. The number of enquiries per page then says enough.
  5. Put the figures side by side every month: analytics tool, inbox and revenue. Look at the direction, not the exact number.

You can read more about measuring and steering on enquiries in our article on GA4 done right and under conversion optimisation.

When this does not apply

If you advertise a lot, with enough traffic to reach Google's thresholds, advanced Consent Mode and enhanced conversions can help your campaigns steer better. That is a trade-off between more measurement data and more data you share, and it should go past a lawyer. Apps use other techniques than cookies, although the rules for consent and personal data stay the same.

If you have few visitors and do not advertise, the question is rather whether you need an analytics tool at all. Counting what comes in is often enough then. Outside the Netherlands the same European rules apply, but the national law and the regulator differ. How we handle data ourselves is in our privacy statement.

What we do not know: how many visitors refuse on average, and how big the gap in your measurement is. That differs per site, so we do not give a figure. We have also not found a statement from the AP on measurements without cookies before consent, such as with advanced Consent Mode.

Frequently asked questions

May I measure visitors without a cookie banner?

Yes, if you only place functional cookies, or limited analytics cookies with no or minor consequences for privacy. For tracking cookies and similar techniques you need consent in advance. If you process personal data, the rules of the GDPR also apply.

Is server-side tagging a way around consent?

No. Server-side tagging gives you more control over which data goes where. But if you set a cookie or read the device in order to track, you still need consent, whichever route the measurement takes.

Does Consent Mode fill the gap in my measurement?

Only partly, and only in its advanced form. Google then estimates the behaviour of people who refused, but only if your site has enough traffic: according to Google, among other things a thousand users a day with consent. With basic, nothing goes to Google without consent.

How do I know how much my measurement misses?

Put the number of enquiries in your analytics tool next to the number that actually arrived in your inbox or records, over the same month. The difference is your gap. That figure is more reliable than an average from someone else's study.

May a website refuse me if I do not accept cookies?

No. According to the AP, you must be able to use a site normally if you refuse tracking cookies. A cookie wall, where refusing means you cannot use the site, is prohibited.

Want to know what your measurement misses, and what is allowed? We look at your banner, your measurement script and your enquiries, and tell you honestly what leaks and what can be better. You can read more about our approach under online marketing. You work directly with Nawid, with a fixed team of five specialists.

Also worth reading: Marketing that pays off in 2026.

Written with the help of AI, reviewed and edited by Nawid Rasa, who is responsible for the content. Sources are listed below the article. Spotted an error? Email hi@empireyo.com.

Get started

Know what your measurement misses?

Let us look at your current setup. You get an honest picture of the leak and the shortest route to a closed chain.

Or message us directly

We usually reply within two working days. Prefer to call or email? That works too. How we handle your details is in our privacy statement.

Sources

Consulted on 4 November 2026.

Empireyo · Zwolle Raising budget while the measurement leaks is carrying water in a sieve. Close the sieve first. Then open the tap.