← Blog·Paid Media

Paid Media

Pixel audit: Meta, Google Ads and TikTok, step by step

A pixel audit checks every ad pixel on your site in one pass. Inventory, firing, values, deduplication, consent and click IDs, platform by platform.

David PombarSwiss army knife at Trackingplan
8 min read · 1760 words

To run a pixel audit for Meta, Google Ads and TikTok, list every pixel on the site, then check five things on each one in the same order: what fires on which page, the values and IDs it sends, deduplication between browser and server events, consent, and the click IDs that tie a conversion to its ad. Do all three platforms in one pass, because the same purchase reaches all of them, and an error in the checkout rarely stays on one.

Most pixel checklists are written for one platform. This one runs across them, with the platform rules side by side, and ends where an audit ends: on one day. What happens between one audit and the next is PPC monitoring. For a deeper check of Meta alone, the Meta Pixel audit goes event by event.

A macro photo of a laptop LCD showing a rainbow, with a magnified inset of the red, green and blue subpixels
From a distance a screen looks fine. A pixel audit is the close-up. Image: Politelyinsulting, Wikimedia Commons, CC BY-SA 3.0

What a pixel audit checks on each platform

The steps are the same everywhere. The names, the keys and the tools change. For the account side of each platform, see the Meta audit, the Google Ads audit and the TikTok audit.

CheckMetaGoogle AdsTikTok
Browser tagMeta PixelGoogle tagTikTok Pixel
Server or first-party dataConversions APIEnhanced conversions (hashed customer data)Events API
Counted once byEvent name and event_id, within 48 hourstransaction_id per conversion actionEvent name and event_id, within 48 hours
Click IDfbclidgclid, from auto-taggingttclid
Browser checkerMeta Ads Data Advisor, Test EventsNetwork tab, a test click on your adTikTok Pixel Helper, ad Preview

LinkedIn and Microsoft Ads follow the same pattern with fewer events. LinkedIn’s help center asks for the tag on the pages of your domain, preferably right before the end of the body in the global footer. Microsoft’s documentation asks for one UET tag on every page, and calls it a prerequisite for conversion tracking. Their account-level checks are in the LinkedIn Ads audit and the Microsoft Ads audit.

The six checks of a pixel audit, in order: inventory, firing, parameters, deduplication, consent and click IDs, then keep the same checks running on real traffic.
The six checks of a pixel audit, in order: inventory, firing, parameters, deduplication, consent and click IDs, then keep the same checks running on real traffic.

Step 1: build the pixel inventory

Start with a list, not a debugger. For every pixel, write down its ID, where it’s loaded (hard-coded, Google Tag Manager, a platform app or plugin) and the pages it loads on.

This is also the first half of a website tracking audit, and it is where the odd findings live:

Mark which pixel each campaign optimizes for. That is the list the next five steps work through.

Step 2: check what fires, and where

Walk the funnel on each platform: page view, product view, add to cart, checkout, purchase or lead. Each action should fire once, on the page that earns it. A purchase that fires again when the thank-you page is reloaded is the classic miss.

The browser checkers help here, within limits:

All three show your visit, in your browser. Note what you saw; Step 6 and the last section come back to it.

Step 3: check parameters and values

A pixel that fires with the wrong value is worse than one that doesn’t fire: the campaign learns from it. On a test order, compare what each platform receives with the order itself.

Step 4: check deduplication between browser and server

When the same conversion travels through the browser pixel and a server API, each platform needs a key to count it once.

The usual failures: an ID generated separately in the browser and on the server, an ID sent on one side only, or a server event renamed during a migration. Each one turns one purchase into two. The patterns are in our post on CAPI deduplication issues.

Run the funnel three times: before answering the banner, after accepting, and after rejecting. For each pixel, note what fires in each state.

Google’s developer docs sets the order: the default consent state goes on every page before any command that sends measurement data. ad_user_data covers sending user data for advertising to Google, and ad_personalization covers personalized ads. With ad_storage denied, no new advertising cookies are set.

Two findings matter most: a pixel that fires before the visitor chooses, and a pixel that keeps firing after a rejection. Both are compliance problems, and both are easy to miss in a browser that accepted the banner months ago. Server events need the same check, because they don’t see the banner unless someone passes the choice along.

Step 6: check click IDs and UTMs

A conversion without its click ID still counts somewhere, but not against the ad that earned it.

Then the UTMs. Open every active ad URL, follow it through shorteners, country redirects and consent walls, and confirm the parameters are still there at the end. Check naming too: facebook, fb and meta are three sources in your analytics.

Keeping it checked after the audit

The audit is now true for the browser you tested, on the day you tested it. The next release can rename an event, drop the transaction ID or load a second pixel, and nothing in the ad platform says so.

That is the part Trackingplan covers. It reads the conversion, UTM and consent signals your real visitors send to Meta, Google, TikTok, Bing and the rest of the 170+ analytics and ad providers it recognizes, browser and server side. It checks Pixel against CAPI against GA4, so the same purchase is counted once with the same value, and it watches the click IDs and event_id that connect a conversion to its ad. When a pixel disappears from a page you send paid traffic to, you get a warning. With your ad accounts connected, it also reads spend, clicks, CPC, conversions and the UTMs and auto-tagging of each ad per campaign, so a pixel problem shows up next to the campaigns and the money it affects. Consent checks, on web and server-side traffic, are part of the Compliance add-on.

It names a cause only when two independent pieces of evidence agree, and if it can’t find one, it says so. It doesn’t touch campaigns, budgets or bids: it checks the signal they run on. Specific checks can be scheduled to run daily or weekly. The pixel checker shows which pixels and CAPI events fire on your real traffic.

FAQ

What is a pixel audit?

A pixel audit is a review of every advertising pixel on a website and of what each one sends: which pixels load on which pages, which events fire, the values and IDs they carry, whether browser and server events are counted once, whether they respect consent, and whether the click IDs that link a conversion to its ad survive the trip to the landing page.

Is a pixel helper extension enough for a pixel audit?

No. Meta Ads Data Advisor and TikTok Pixel Helper are good for step two, checking what fires on a page. They run in your browser, on your visit, with your consent choice. They don’t show server events, other browsers, or visitors who answered the banner differently.

How often should I run a pixel audit?

Run a full one when you take over an account, before a site migration or a checkout change, and after a new consent banner or a server-side setup goes live. Between audits, watch the same checks on real traffic, because one release can change what a pixel sends.

What is the difference between a pixel audit and a website tracking audit?

A website tracking audit covers everything the site sends, analytics included. A pixel audit is the part that feeds the ad platforms: Meta, Google Ads, TikTok, LinkedIn, Microsoft Ads and the rest. It is the part where an error changes what bidding learns.

David PombarSwiss army knife at Trackingplan

Read more from David, a Senior Product Strategist with 18+ years in digital product development and an atypical error detection knack.

Read next

All Blog →

Trackingplan

See everything. Miss nothing.

Your implementations audited around the clock with real-time, real-user data. Real-time alerts about errors or changes in your data, campaigns, pixels, privacy and consent. Let AI flag issues before they cost you.