Analytics and Tracking

PII

Also called Personally identifiable information

Any data that can identify a specific person, which Google and Meta forbid you from sending to their measurement tools.

Quick facts: PII

Category
Analytics and Tracking
Also called
Personally identifiable information
Level
Intermediate
Affects
Analytics account standing, legal exposure, report sharing
Where to see it
GA4 page path and search reports, Google Tag Manager preview, browser network tab
In this article4
  1. How PII works in analytics
  2. Why PII matters
  3. Where PII goes wrong
  4. How to act on it

How PII works in analytics

Definitions vary between laws, but for measurement work the practical test is simple: could this value, alone or joined with something else you hold, point at one named human? Names, email addresses, phone numbers, postal addresses, government identity numbers and account numbers all qualify. So do combinations that look harmless individually.

What matters day to day is that Google’s terms prohibit sending it to Analytics or Google Ads, and Meta’s terms say the same about its own tools. This is not a setting you can switch off. It is a condition of using the product, and the platforms scan for it.

Why PII matters

Because the damage is wholesale rather than retail. A single hit carrying an email address in the page path does not spoil one row; it puts the property in breach, and the remedies available to Google include deleting data or suspending access. There is no way to surgically remove the offending values from data already collected, so a leak discovered late can cost you the historical reporting you were relying on.

There is a legal layer on top. Personal data sent to a platform under terms that forbid it is processing you cannot properly document, and it will not appear in your privacy notice because nobody intended it. And a practical layer underneath: a report full of email addresses in the page report cannot be shared with a client, an agency or a junior analyst without spreading the problem.

Where PII goes wrong

Almost always by accident, and almost always through the URL. A form that submits by GET puts every field into the query string, so the email address lands in the page location of the next page view. Booking systems, password reset flows and account areas put references, tokens and sometimes addresses straight into the path. Chat widgets and survey tools add their own parameters.

The other regular offender is a user ID set to the customer’s email address, on the reasonable-sounding logic that it is unique and stable. It should be an internal identifier that means nothing outside your own database. Worth separating from this: enhanced conversions deliberately send customer data through a sanctioned, hashed channel, which is a different mechanism from pasting an address into an event parameter.

How to act on it

Search your page path and page location reports for an at sign and for long strings of digits, then look at your search terms and event parameters for the same. If you find nothing, check again after every new form or booking tool goes live.

Fix the source first: forms should submit by POST, and confirmation pages should carry a reference that only your systems can resolve. Then add a safety net in your tag manager that strips known parameters and rewrites the page path before anything is sent — redaction after collection is not available to you, so it has to happen before the hit leaves the browser. If you are unsure what your tags are transmitting, a tracking audit is the quickest way to find out.

Do and do not

Do

  • Search page paths for at signs and long digit strings
  • Set the user ID to an internal pseudonymous identifier
  • Strip risky parameters in the tag manager before hits send

Do not

  • Let forms submit by GET into a confirmation URL
  • Put an email address into any event parameter
  • Assume you can delete individual values after collection

Questions people ask about this

Is an IP address personally identifiable information?

Under European law it is generally treated as personal data, because it can identify a household or person when combined with other records. Google's own policies treat it separately from the identifiers it forbids you to send, and its tools handle it internally. The safe working position is to treat an IP address as personal data in your privacy notice and retention decisions.

What happens if I accidentally send email addresses to Google Analytics?

Stop the source immediately, then assess how far back it goes. Google's terms allow it to delete affected data or restrict the account, and there is no way to remove individual values from data already collected, so the usual outcome is deleting the affected date range or property. Document what happened, because it is a personal data incident as well as a policy breach.

Can I use a customer's email address as the user ID?

No. The user ID should be a pseudonymous internal identifier that means nothing to anyone outside your own systems, such as a database key or a generated string. Keep the mapping between that identifier and the real person in your own database, where it is covered by your privacy notice and can be deleted on request.

Related terms

Found this useful?

Share it, or ask an AI to summarise it

Back to the glossary

Knowing the term is the easy part

Applying it to your own site and budget is the work. Book a call and I will tell you what actually applies to you.