Public pages use no advertising trackers. The secure portal uses an essential session cookie and the service providers described in our privacy policy.

Skip to main content
← Back to Resources
Process·November 10, 2026·7 min read

Checking a Case File Against Your Tag Manager

How a web agency can match each request in a website privacy record to the tag, trigger and container version that sent it, and write up the result for counsel.

PR
Privacy Resolve
Last reviewed November 10, 2026

When a client receives a website privacy letter, its web agency is usually asked one question first: did our setup send this? The case file records what the visitor's browser sent on the visit date. The tag manager's history records why. This guide shows how to put the two side by side, request by request. For a plain-language tour of the case file itself, start with How to Read Your Evidence File.

What can a tag manager confirm about a case file?

It can confirm which tags were published on the visit date, what triggered each one, which consent checks applied, and where a server container sent data, because each published version is kept with its date and publisher.

Google's documentation for Google Tag Manager describes the record. "A version is a snapshot of a container configuration at a particular time." "Every time a container is published, a version of that container is recorded." And "Tag Manager maintains a publish history, so you can see when versions were live and who published them."

A tag manager cannot confirm tags that never passed through it: code pasted into the site's theme or header, plugins and platform integrations, and widgets that load their own vendors. Those need the site's own code history.

The visit date is what matters. The federal Wiretap Act gives the civil claim to "any person whose wire, oral, or electronic communication is intercepted, disclosed, or intentionally used in violation of this chapter," 18 U.S.C. § 2520(a), so the question is what the configuration did on that date, not what it does today.

Which container version was live on the visit date?

The most recent version published before the visit time; open Versions in Google Tag Manager, read the Published column, and note the version number, its publish time and who published it.

Google's instruction is direct: "To see the publish history, go to Versions and look for entries with a date in the Published column." Put both clocks in one time zone before comparing. Check the case file's stated time zone; the evidence files described in How to Read Your Evidence File use UTC unless they say otherwise.

Check environments too. Every container has a Live environment, and Google says that "[w]hen you publish a custom environment, you will not send those changes to the live version of your container." A version published only to a staging environment did not run for ordinary visitors.

Export the version that was live and keep the file. Google says "Google Tag Manager containers can be exported as JSON": in Admin, "click Export Container," then "[s]elect the desired container or workspace version." The export is a text record of the version's tags, triggers, variables and consent settings, and Google notes it suits workflows that use "text-based tools such as diff and Git," so two versions can be compared line by line.

How do you match a request in the record to a tag?

Match the destination and the ID: each request names the vendor's host and usually carries the tag's own ID, such as a G- measurement ID or a Meta Pixel ID.

  1. List the third-party requests. For each, note the host, the time, the event name and the page address it carried. How to Read a Website Data Transmission explains those fields.
  2. Find the ID. Google describes a GA4 measurement ID as "'G-' followed by a combination of numbers and letters." Meta's documentation lists the "Pixel ID and the Facebook Cookie" as pixel-specific data the pixel collects. Other vendors use their own account or tag IDs.
  3. Search the exported container for that ID. If it appears, you have the tag: note its name, type and firing triggers. If it does not, the request came from something the container does not control, such as code in the theme or header, a plugin or platform integration, a third-party widget, a second container (search the page source for other GTM- IDs), or a server container.
  4. Match the event. Some events come from a tag's own settings rather than from a trigger anyone wrote. Meta's documentation says "[e]ach time the Pixel loads, it automatically calls fbq('track', 'PageView')," and Google's enhanced measurement adds events such as view_search_results, form_start and file_download when it is on. Site Search and Form Events covers those.

How do you tell what made a tag fire?

Read the tag's firing triggers in the exported version, then reproduce the visit in preview mode, which shows which tags fired, in what order, and what triggered each one or kept it from firing.

Google says a "previewed site is connected to Tag Assistant so that you can inspect which tags fired and in which order." Preview runs the container "as if the current container draft was deployed," and it "is not visible to your regular website visitors." Because preview runs the current draft, not the visit-date version, a faithful test needs that version published to a custom environment such as staging and viewed on the staging site or through that environment's preview link. Publishing to a custom environment does not change the live container. Note in writing that you did it and when.

Run the test on a test copy of the site, use made-up words in every search box and form, and never submit a real intake, appointment or contact request.

Findings that come up often:

  • Triggers set to fire on all pages with no consent condition.
  • Custom HTML tags that load a vendor's base code; search the export for fbq( and gtag(.
  • Triggers keyed to form submissions or clicks on booking buttons.
  • Page groups that include condition, provider, booking or search results pages.

They show whether each tag was set to wait for consent: Google's Consent Overview lists consent settings across the container, and a tag with no additional consent settings performs no additional checks.

Google describes the screen: "The Consent Overview page provides a high level view of all the consent settings across the tags in your container." Some tags have built-in checks, and "[t]ags with built-in consent checks include logic that changes the tag's execution behavior." A tag set to require additional consent "will only fire if the status of all specified consent types are 'granted' when the tag is triggered." For a tag without them, Google's wording is "[n]o additional consent checks are performed by this tag."

A built-in check changes behavior, but not always to silence. Google's documentation says that in the advanced consent mode setup, while consent is denied, "the Google tags send measurements without cookies," and that with advertising storage denied, "[f]ull page URLs are collected, including ad-click information in URL parameters." So a Google request after a refusal may be advanced consent mode working as Google documents it rather than a broken trigger; the request's consent parameters show which. Whether sending it was right is a separate question. See Google Consent Mode: Basic Versus Advanced.

The banner itself usually lives in a separate consent platform. Export its configuration for the visit date as well, including what each button was set to do.

What if a request went to the site's own domain?

Then look for a server container or another first-party collector: in server-side tagging, the browser sends its requests to a server the site controls, which can forward the data to vendors.

Google describes the server container as one that "runs on a server that you control." It "receives web requests from the user's device and transforms those requests into events," and "only you have access to the data in the server until you choose to send it elsewhere." The server container has its own versions, tags and logs. Export the version live on the visit date and list each tag that forwarded data, with its destination. Meta Conversions API and Server-Side Tagging explains why the vendor is still the recipient.

What should stay untouched while a claim is open?

Keep every container version, the consent platform's history, server logs and the old site code, and publish fixes as new versions instead of deleting or overwriting what was live.

Fixing the site is right, and it changes future visits, not the recorded one. Write down each change, the version number it went out in, and who approved it. Do not delete old versions or logs to make the history look cleaner. What to Do After a Privacy Demand Letter covers the client's side of keeping records.

What should the agency give the client and its counsel?

A short written match: for each request in the record, the tag that sent it, the container version and trigger, the consent setting, and any request you could not match, with the exported files attached.

A sample entry, with illustrative values rather than a record of a real visit:

  • Request: 14:02:11 UTC, www.facebook.com/tr, event PageView, page https://example.test/services/knee-pain.
  • Tag: "Meta base code," a Custom HTML tag in container version 41, trigger "All Pages," no additional consent settings.
  • Notes: fired before the banner choice; removed in version 44, published on a stated date by a named reviewer.

If a request in the record could not have come from anything the client runs, say so with the evidence, and use the factual correction route. The client's attorney may ask you to prepare the write-up at their direction; follow their instructions on format and delivery.

What sources does this guide rely on?

This guide relies on the vendor documentation and statute listed here, quoted from the pages linked. Checked on November 10, 2026.

General information only, not legal advice, and not a substitute for the advice of an attorney. Laws, deadlines, defenses and outcomes vary by jurisdiction and facts.