Server side tracking is not automatically GDPR compliant. Moving tags to your own server, or switching affiliate conversions to postbacks, changes where data flows, but if anything still reads from or writes to the visitor's device (a cookie, local storage, a pixel, a click ID in a URL, a fingerprint), the ePrivacy consent rules still apply, and the GDPR still governs every piece of personal data you pass on. Done well, server-side tracking can make compliance easier; done as a way to route around a consent banner, it makes things worse.
This guide explains where the legal lines fall in the EU and UK, what the EDPB has said about the technical scope of the cookie rule, and how to set up server-side tagging and S2S tracking so that it holds up to scrutiny. It is general information, not legal advice; take professional advice for decisions about your own set-up.
Why server side tracking and GDPR still go together
Two separate sets of rules apply to tracking, and server-side set-ups affect them differently.
- The cookie rule (Article 5(3) of the ePrivacy Directive in the EU; regulation 6 of PECR in the UK) covers storing information on, or gaining access to information from, a user's device. It applies to any information, personal or not, and the default is prior consent unless an exemption applies.
- The GDPR (or UK GDPR) covers what happens to personal data afterwards: lawful basis, transparency, minimisation, sharing with third parties, international transfers and security.
Server-side tagging usually still sets a first-party cookie and still runs JavaScript in the browser to send events to your tagging server. That is storage and access on the device, so the cookie rule is engaged exactly as before. What changes is the second half: your server receives the data first and decides what to forward to analytics, ad platforms or affiliate networks. That control point is the genuine privacy benefit, but only if you use it.
Common mistake: treating "first-party" and "server-side" as words that remove the consent requirement. Neither the EU nor the UK rules exempt a technology because it is hosted on your domain. The test is what the technology does on the device and why.
What the EDPB says about the technical scope of Article 5(3)
The European Data Protection Board's Guidelines 2/2023 on the technical scope of Article 5(3) (final version October 2024) were written precisely to close the "it isn't a cookie" argument. The key points for webmasters:
- Gaining access is broad. It covers instructing the device to send back information, whether through cookies attached to HTTP requests, JavaScript that makes calls, or SDKs. It does not matter that a different party receives the data.
- Pixels and tracking links are in scope. Tracking pixels and links carrying identifiers are at least temporarily stored on the device and cause it to send data back.
- Unique identifiers are in scope. A persistent ID derived from personal data, such as a hashed email address that client-side code collects and sends, counts as gaining access.
- IP-based tracking can be in scope where the IP address originates from the user's terminal equipment, though the analysis is case by case.
The guidelines do not say every one of these needs consent. They say Article 5(3) applies, so you then need either consent or a genuine exemption (strictly necessary for a service the user asked for, or transmission of a communication).
Fingerprinting and ePrivacy
Fingerprinting (combining browser, device and network attributes to recognise a visitor without a cookie) is squarely within the cookie rule. The EDPB guidelines confirm that collecting context data such as HTTP headers for this purpose can trigger Article 5(3), and the UK ICO's storage and access technologies guidance treats fingerprinting the same way as cookies. Fingerprinting for advertising or cross-site tracking needs consent. Using it to dodge a refusal is the kind of practice regulators single out, because the user has no practical way to clear it.
Postback tracking privacy: is S2S tracking GDPR-friendly?
Affiliate postback tracking is often described as "cookieless", and in one sense it is: the conversion is reported server to server from the advertiser to the network or your tracker, rather than by a pixel firing in the user's browser. But a typical S2S flow still involves the device:
- The visitor clicks your link, which carries a click ID or sub-ID in the URL.
- The advertiser's landing page reads that click ID and usually stores it in a cookie or local storage so it can be attached to a later sign-up or purchase.
- When the user converts, the advertiser's server sends the click ID back to the network's postback URL.
Step 2 is storage on the device, and the click ID is an identifier tied to an individual's journey. Whether that storage needs consent is a live question; we cover the arguments in do affiliate cookies need consent?. The practical position for most European operators is that cross-site attribution for commission purposes is not "strictly necessary" from the user's point of view, so the cautious approach is to cover it in your consent mechanism or to be able to justify, in writing, why an exemption applies.
On the GDPR side, S2S tracking can be more privacy-friendly than pixels, because you control exactly which fields are posted back. A postback that carries only a click ID, a payout and a status is very different from one that forwards an email address or full IP. For the mechanics, see our comparison of S2S and pixel tracking.
Server side GTM and consent: how it should work
Server-side Google Tag Manager and similar tagging servers sit between the browser and vendors. Consent still has to be collected in the browser, and the consent state has to travel with each event so the server can act on it. A sound design looks like this:
| Layer | What it should do | What to check |
|---|---|---|
| Consent banner or CMP | Collect a choice before non-exempt storage or access happens | Equal accept and reject options; nothing fires before a choice |
| Browser tag | Send events to your tagging server with the consent state attached | No vendor scripts load directly in the browser for refused purposes |
| Tagging server | Route each event only to vendors the user agreed to; strip or truncate fields | Rules per vendor; IP truncation; no user IDs to ad platforms without consent |
| Vendors | Receive only what you forward, under a contract | Processor terms or controller-to-controller terms; transfer safeguards |
If you use Google's consent signals, read our explainer on Google Consent Mode v2, and our technical guide to server-side tagging for the build itself.
UK and EU differences in 2026
The UK has diverged. Since 5 February 2026, the Data (Use and Access) Act 2025 has amended PECR so that storage and access purely for statistical purposes (analytics to understand and improve your own service) and for some functionality purposes no longer need consent, provided you give clear information and a simple way to object. That exemption is narrow: it does not cover advertising, profiling or cross-site tracking, and the ICO expects any analytics provider to act only as your processor. A server-side analytics set-up that keeps data aggregated and first-party may therefore fit the UK exemption where it would still need consent in most EU countries. See analytics without cookie consent for the detail.
In the EU, nothing has changed yet. The Commission's Digital Omnibus proposals would add some consent exemptions, but they are not law. Current national ePrivacy rules and the GDPR apply.
A compliance checklist for server-side and S2S tracking
- Map every device interaction. List each cookie, storage key, pixel, script and URL parameter the set-up uses, including on the tagging subdomain.
- Assign a purpose and a legal route to each. Strictly necessary, UK statistical exemption, or consent. Write it down.
- Pass consent state server-side. Your server must know what the user agreed to and route accordingly.
- Minimise what leaves your server. Truncate IPs, drop user agents you don't need, avoid hashed emails unless the user consented to that sharing.
- Settle the roles. Is each vendor your processor or an independent controller? Our guide to controller vs processor roles covers affiliate networks and trackers.
- Check transfers. A tagging server hosted in the EU that forwards to a US vendor is still a transfer.
- Disclose it. Your privacy and cookie notices should describe server-side processing in plain words, not hide it behind the word "first-party".
- Test refusal. Reject everything, then inspect network requests and server logs to confirm nothing non-exempt is set or forwarded.
Frequently asked questions
Does server-side tracking remove the need for a cookie banner?
No. If the set-up stores or reads anything on the device for non-exempt purposes, you still need consent. Server-side tagging changes where data goes after collection, not whether the collection step needs consent.
Can I use fingerprinting if a user rejects cookies?
Not for tracking purposes. Regulators treat fingerprinting as storage or access on the device, and using it to override a refusal undermines the consent itself.
Are postbacks personal data?
Often, yes. A click ID that can be linked back to a specific visit or conversion is pseudonymous personal data in the hands of whoever can make that link. Keep the payload minimal and cover it in your contracts.
Next steps
Start by auditing what your current tracking actually sets and sends, then compare platforms in our affiliate tracking platforms directory for ones that support consent-aware routing and minimal postbacks. For the bigger picture, our GDPR and cookies hub links every guide in this series.