November 2026 · 9 min read
GDPR Link Tracking: How to Count Clicks Without Tracking People (2026)
Counting clicks on a link is not banned under the GDPR. What creates obligations is the data each click carries with it. An IP address, a cookie ID or a link made for one person can turn a click record into personal data, and tracking links can also fall under the EU ePrivacy rule on reading data from a device.
This guide to GDPR link tracking covers what the law and the European Data Protection Board (EDPB) say, which click fields are the risky ones, and a setup that keeps your campaign numbers while collecting less. Each legal point links to an official EU source, checked on 19 November 2026.
We make TrimLink, a link shortener, so we also list exactly what it stores for each click, including the parts that touch personal data. This is not legal advice. If your traffic is regulated or your setup is unusual, ask a lawyer or your data protection authority.
Two sets of rules apply to link tracking
The GDPR: is the click record personal data?
The GDPR applies to personal data, which Article 4(1) defines as any information about a person who is identified or identifiable. It names online identifiers as one way a person becomes identifiable. Recital 30 goes further. It lists IP addresses and cookie identifiers as online identifiers that leave traces, which can be used to build profiles of people and identify them.
The EU Court of Justice made the IP point concrete in Breyer, Case C-582/14, decided on 19 October 2016. It ruled that a dynamic IP address logged by a website operator can be personal data for that operator when it has legal means to identify the person with extra data held by the internet provider. That ruling applied Directive 95/46, which the GDPR replaced, and the GDPR's definition still turns on whether a person is identifiable.
On the other side, Recital 26 says data protection principles do not apply to anonymous information. That sounds like an easy way out, but a record with no name can still identify someone when combined with other data. Leaving out names is not the same as being anonymous.
If click data is personal data, two GDPR rules matter most for link tracking. Article 5(1)(c), data minimisation, says data must be adequate, relevant and limited to what is necessary for the purpose. Article 6 requires a legal basis, and for analytics many businesses rely on legitimate interests under Article 6(1)(f). That basis works only when your interest is not overridden by the interests, rights and freedoms of the people clicking.
ePrivacy: does the tracking reach into the device?
The second rule is Article 5(3) of the ePrivacy Directive. It covers storing information on a user's device, or reading information already there, and allows it only with consent or under narrow exemptions, such as when it is strictly necessary for a service the user asked for. It is the rule behind cookie banners, and it applies whether or not the information is personal data.
In Guidelines 2/2023 on the technical scope of Article 5(3), version 2.0 adopted on 7 October 2024, the EDPB says this rule is not limited to cookies. It covers several techniques that matter for links:
- Tracking links. An identifier added to a URL that plays no part in loading the page makes it a tracking link (paragraph 49). For links sent by email or text message, the EDPB writes that "Article 5(3) ePD is applicable, even if this storage is not permanent" (paragraph 50).
- Fingerprinting from headers. Using HTTP headers such as the user agent to fingerprint a device can bring Article 5(3) into play (paragraph 43).
- Tracking by IP only. Following someone's browsing with nothing but their IP address, sometimes across sites, can also fall under it (paragraphs 54 and 55).
The EDPB also notes that when Article 5(3) applies, consent is not automatically required, and each case still has to be checked against the exemptions (paragraph 56). National data protection authorities enforce these rules and publish their own guidance on exemptions, for example for audience measurement, so check the guidance in the countries you target.
Which click data counts as personal data
Here is how common click fields line up against those rules. A risk here means the field can identify a person or bring in the device rule above, not that collecting it is always unlawful.
| Click field | Why it is a risk | Lower-data option |
|---|---|---|
| IP address | An online identifier under Recital 30 and Breyer | Read a coarse location from it, then don't store it |
| Cookie or visitor ID | Links one person's visits together | Set no cookie on the redirect |
| Full user agent string | A fingerprinting input under paragraph 43 | Keep browser family, OS family and device type |
| An ID for each recipient in the URL | A tracking link under paragraphs 49 to 51 | Use one link per placement |
| Referrer with its query string | Query strings can hold email addresses, tokens or names | Keep the referring site and path only |
| Free-text UTM values | People paste email addresses and customer IDs into them | Accept campaign names only |
For comparison, Google's own help page says Google Analytics 4 sets a _ga cookie that lasts 2 years and is there to tell users apart. That is the kind of ID the second row is about.
Per-person links are the biggest risk
Most of what a marketer wants from link tracking is about channels. Did the newsletter beat the Instagram bio? Did anyone scan the poster? None of that needs to know who clicked.
The problem starts when every recipient gets their own link. An email tool that reports which subscriber clicked can only do it by giving each recipient a link that identifies them. Under paragraphs 49 to 51 of the EDPB guidelines, that is the tracking link the ePrivacy rule covers, and the click history it builds is personal data under the GDPR as well.
A link per placement avoids that. One short link for the November newsletter, one for the bio, one for the printed flyer. Every reader of the newsletter gets the same link, so a click tells you the newsletter worked and nothing about the person. Our Black Friday tracking playbook shows that setup with a naming scheme you can copy.
A link tracking setup that collects less
- Make one short link per placement, not per person. Name it after the channel and the campaign, never after a customer.
- Tag the destination with campaign UTMs. Use values like
newsletter,emailandbf-2026. Never put an email address, a customer number or a name in a UTM value. Our UTM parameter guide covers the five tags, and the UTM builder writes them for you. - Strip query strings from referrers before you store them. The site and the path are enough to see where clicks came from.
- Reduce the user agent to families. Device type, browser family and OS family answer the reporting questions a campaign has.
- Set no cookie on the redirect. A redirect has no reason to recognise a visitor the next time.
- Choose a retention period and write it down. Data minimisation applies to time too. Decide how long you need click history, then delete it or keep only totals.
- Mention link tracking in your privacy notice. Say which links are tracked, what is counted and why.
- Check where your shortener stores data. If it is outside the EEA, look for the transfer safeguard it relies on.
For reading the numbers once they come in, see how to track link clicks and the link tracking guide.
What TrimLink stores for each click
We built TrimLink to count clicks without identifying the people who click. Here is exactly how that works, and you can check each point against our privacy policy.
Stored with each click:
- The time of the click.
- Country, region and city, read from location headers our hosting provider adds to the request. The IP address is not stored with the click.
- Device type, browser family and OS family, with no version numbers.
- The referring page with its query string and fragment removed.
- Whether the click came directly, from a QR code scan or through our API.
- The five standard UTM values, kept only when they look like a campaign name. Values that look like an email address, a phone number or a URL are dropped, and each value is capped at 100 characters.
Not stored with a click: the IP address, the full user agent string, a cookie or any ID that links two clicks to the same person. The redirect sets no cookie for the visitor. Clicks from bots and link-preview crawlers are not counted.
The part that does touch personal data. To stop one device from inflating a count, we keep a counter for about 6 minutes after a click. It is keyed to a SHA-256 hash of a few request headers (user agent, language, accepted formats and browser hints when sent), or to the IP address when those headers are missing. It lives only in our cache and never reaches the click record. Under paragraph 43 of the EDPB guidelines, a header-based check like this can count as reading information from the device, so we name it here instead of leaving it out.
A few more facts you should know:
- If a link is paused, broken or expired, the visitor sees a TrimLink page instead of a redirect, and that page loads our error monitoring.
- Our database and cache run in the United States. For data from the EEA, the UK and Switzerland, our privacy policy says we rely on our providers' data processing terms, which include the European Commission's Standard Contractual Clauses.
- Our privacy policy names legitimate interests as the legal basis for showing link owners their click statistics.
We don't call the click record anonymous. It holds no identifier, but city, time and device together can still single someone out in a small audience, so treat it as data you have to account for. None of this makes a setup compliant on its own. That depends on your purpose, your notice and the rest of your stack. For the wider privacy case, read URL shortener privacy.
Frequently asked questions
Is an IP address personal data under the GDPR?
It can be. Recital 30 names IP addresses as online identifiers, and in Breyer (Case C-582/14) the EU Court of Justice held that a dynamic IP address can be personal data for a website operator with legal means to identify the person through the internet provider. Treat IP addresses as personal data unless you have a strong reason not to.
Do I need consent to track link clicks?
It depends on what the tracking does. The EDPB says tracking links in emails and text messages fall under Article 5(3) of the ePrivacy Directive, and also that this does not automatically mean consent is needed (paragraph 56). A link per placement with no per-person ID and no cookie collects far less than a link made for each recipient. Ask your national data protection authority or a lawyer about your own case.
Are UTM parameters personal data?
Campaign values like newsletter or spring-sale describe a campaign, not a person, so on their own they usually aren't. They become personal data when someone puts an email address, a customer ID or a name in them. Keep UTM values to channel and campaign names.
Can I see which subscriber clicked a link in my newsletter?
An email tool can, by giving each subscriber a unique link. That is a per-person tracking link in the EDPB's sense, and it creates personal data. If you only need to know whether the newsletter worked, use one short link per issue and count clicks for the whole send.
Part of the complete guide to link management.
Start Free — No Credit Card
Shorten, brand, and measure every link. Track every click, track zero people.
Keep reading
More on links, tracking, and keeping your data yours.