PII compliance is collecting, storing, sharing, and deleting personally identifiable information the way the laws that apply to you require. In practice: know which personal data you collect, keep only what you need, protect it, send it only where the person agreed, and be able to prove all of it.
Most PII compliance programs cover databases, CRMs, and access rights. Fewer cover the requests your website and apps send to Google, Meta, and TikTok every second, which is where a lot of personal data actually leaves the company. This post is the compliance side of our guide to privacy compliance for analytics: what the laws ask, and a checklist that includes the tags.
What counts as PII, and why the label matters
PII is any data that identifies a person, on its own or combined with other data. Names, email addresses, phone numbers, and government IDs identify someone directly. IP addresses, device IDs, cookie IDs, and precise location identify someone once they are combined. Our PII definition and examples has the full list.
Two distinctions shape the work:
- Sensitive vs non-sensitive. A Social Security number, a card number, a diagnosis, or a religion can cause real harm when exposed, so laws put stricter rules on them. A city or a job title is personal, but low risk alone.
- PII vs personal data. “PII” is the US term. The GDPR uses “personal data” and defines it more widely, including “an online identifier” (GDPR, Article 4). A cookie ID that a US team would not call PII is personal data in Europe.
If you serve visitors in both markets, work with the wider definition. It saves arguing field by field.
PII compliance requirements by regulation
There is no single PII law. The requirements come from the laws that apply to your users and your sector.
| Regulation | Where | What it asks of tracking | What it costs |
|---|---|---|---|
| GDPR | EU and EEA | A lawful basis for every use of personal data, data minimization, and breach notification within 72 hours (Article 33) | Up to 20 million€ or 4% of global turnover, whichever is higher (Article 83) |
| ePrivacy Directive | EU | Consent before storing or reading anything on the device that is not strictly necessary, which covers most analytics and ad cookies (Article 5(3)) | Enforced by national authorities, often together with the GDPR |
| CCPA / CPRA | California | Notice, opt-out of “sale” and “sharing”, and honoring the Global Privacy Control signal. Sharing visitor data with ad and analytics trackers can count as a sale: that was Sephora’s $1.2 million settlement in 2022 (Faegre Drinker) | Up to $7,988 per intentional violation (Privacy Law Map) |
| Other US state laws | 20 states with a comprehensive privacy law; California, Colorado, Connecticut, and New Jersey require honoring GPC from 2026 (Multistate) | Opt-outs, sensitive-data consent, and data protection assessments, with differences by state | Set by each state attorney general |
| HIPAA and FTC rules | US health | Health information that a pixel sends to a vendor from a patient portal or a login page can be an impermissible disclosure (HHS guidance on online tracking) | The FTC settled with GoodRx for $1.5 million and with BetterHelp for $7.8 million over health data sent through pixels (DWT) |
Read the table in one direction: the laws differ in thresholds, but every one of them asks you to know what data leaves, where it goes, and on what basis. That is a question your tags answer, not your privacy policy.
The PII compliance checklist
Seven steps, in order. The first five set the rules. The last two keep them true after the next release.
1. Map where PII enters and where it goes
List every place a person types or is identified: sign-up, checkout, contact forms, search, login, and app screens. Then follow each field. Does it go into the dataLayer? Into a URL? Into an event property? Which vendors receive that event? The map is not finished until each field has a list of destinations.
2. Classify each field by sensitivity
Tag every field on the map as sensitive, personal, or general. Sensitive data (health, finance, government IDs, precise location) gets the strictest rules and, in several US states, its own consent requirement. The PII examples by category help when a field is borderline.
3. Write down which data may go to which vendor
This is your PII policy, and it can be one table: field, purpose, allowed destinations, legal basis, retention. Data minimization lives here. If a vendor does not need the email to do its job, it does not get the email, hashed or not.
4. Keep PII out of URLs, page titles, and event properties
Google Analytics forbids sending PII and lists the usual offenders: URL paths and parameters, page titles, campaign parameters, site search terms, and custom dimensions (Google). Use POST for forms, keep identifiers out of query strings, and push only the fields each tag needs to the dataLayer. When a platform requires a personal identifier, as Meta’s and Google’s enhanced matching do, send it hashed and only after consent.
5. Tie every tag to consent
Each tag that reads or sends personal data should fire only for the consent category it belongs to. Configure the CMP, set up Google Consent Mode v2, and check that the vendors you declare in the banner are the ones that load. A banner that says “reject” while the pixels still fire is a classic cookie fine: the CNIL fined Condé Nast 750,000€ in 2025 for cookies set before the banner was answered and a “refuse all” that did not work (noyb).
6. Monitor the data that actually leaves, continuously
Steps 1 to 5 describe the site on the day you audited it. Then a release renames a form field, a marketer adds a pixel, or a GTM change copies the whole user object into an event. None of that shows up in the policy. It shows up in the hits. Checking real traffic on an ongoing basis is the only way to know the rules still hold.
7. Keep evidence and a response plan
When something leaks, you need to answer three questions fast: what was sent, to whom, and since when. Keep a record of findings and fixes, know how to request deletion from each vendor (GA4 has data deletion requests), and know who decides whether it is a notifiable breach. Under the GDPR, the clock is 72 hours from when you become aware.
Where analytics breaks PII compliance
Most PII leaks in analytics are not attacks. They are ordinary implementation details. Three common paths:
- The URL. A confirmation page at
/order/confirm?email=ana%40example.comsends that address to every tag that records the page location. GA4, the Meta pixel, and most ad tags do. - The dataLayer. A developer pushes the whole
userobject for convenience. A GTM variable readsuserinstead ofuser.id, and the name, email, and phone ride along to every tag that uses it. - The form field captured as an event. A “form submitted” event takes the search or message text as a property. Visitors type their phone number into it.
This is what the first case looks like in a GA4 request, decoded:
https://region1.google-analytics.com/g/collect
?v=2&tid=G-XXXXXXX&en=page_view
&dl=https://shop.example.com/order/confirm?email=ana%40example.com
The privacy policy is correct. The CMP is configured. The email still reaches Google, and from there every report and export that includes the page.
How Trackingplan catches PII before legal does
Trackingplan inspects every hit for PII leaks, Consent Mode misfires, and vendors firing without permission, across web, mobile, and server-side. It reads the requests your real visitors already send, so it sees the leak that appears after the audit: emails, names, IDs, and addresses traveling in parameters, page paths, or referrers, to any platform.
Personal fields are masked on the device, so the data never reaches Trackingplan. Each finding comes with the field, the vendor that received it, when it started, and how many hits it affects, so the fix is a tag change, not an investigation. The mechanism, field by field, is in how PII leak detection works.

No tool makes a company compliant by itself. What monitoring gives you is the evidence for steps 6 and 7: what your tags sent, to whom, and under which consent state.
FAQ
What does PII compliance mean?
Handling personally identifiable information the way the applicable laws require: collect only what you need, protect it, share it only on a valid legal basis, honor people’s rights, and keep proof that you did.
What are the main PII compliance requirements?
Know what personal data you hold and where it goes, minimize it, secure it, get consent where the law requires it, honor access, deletion, and opt-out requests, and notify breaches on time. The exact rules depend on the GDPR, the US state laws, and sector laws such as HIPAA.
Is a hashed email still PII?
Under the GDPR, yes. Hashing is pseudonymization, and pseudonymized data that can be linked back to a person is still personal data (Recital 26). Hash before sending, but treat the hash with the same consent rules as the email.
Who is responsible for PII compliance?
Legally, the company that decides why and how the data is processed. In practice, it is shared: legal and the DPO set the rules, and the analytics, marketing, and development teams decide what the tags actually send.
How do I check whether my website leaks PII?
Open the browser’s network panel, go through your forms and confirmation pages, and search the outgoing requests for the email and phone you entered. That answers it for one session. To know it stays true after every release, monitor real traffic continuously.