To check that the Meta Pixel fires on your landing page, install Meta Ads Data Advisor, the browser extension Meta used to call Meta Pixel Helper, open the landing page and read the badge on its icon: it should show one Pixel and a PageView. Then submit the form and confirm in Test Events, in Events Manager, that Meta received one Lead. The extension proves the script ran, Test Events proves the event arrived.
The two tools, and what each one sees
They answer different questions, and a test that uses only one of them misses half the failures.
| Tool | Where it lives | What it tells you |
|---|---|---|
| Meta Ads Data Advisor | an extension for Chrome and Microsoft Edge, with a badge on the icon | how many Pixels loaded on the page you are viewing, which events fired and when, plus warnings and errors |
| Test Events | Events Manager, under Data Sources and then your Pixel | whether Meta received the events, and what the payload carried |
Meta describes Data Advisor as formerly Meta Pixel Helper, which is why older tutorials use the other name. Installs that existed before the rename were updated automatically, so there is nothing to migrate.
A test run that catches most setup errors
- Turn off your ad blocker for the landing page, or allow it for the domain. Meta’s own testing steps for Data Advisor include deactivating the blocker, and a blocked script looks exactly like a missing one.
- Open the landing page. The badge should show one Pixel, with a PageView already listed.
- Submit the form with a deliberate mistake, such as letters in the phone number. No Lead should fire.
- Submit it properly. One Lead.
- Refresh the confirmation screen. Still one Lead, not two.
- Look for an event marked Auto-detected next to the ones you coded.
- Run the server side with a test event code and check that the two copies deduplicate.
That sequence tests the two failures that cost the most: an event that fires when it should not, and a second copy that Meta cannot pair with the first.
When the badge shows no Pixel at all
The script never ran, and the cause is usually on the page rather than in Meta.
- The base code is missing from that template. Landing page tools often have a separate setting per template, so one variant can be tracked and the next one not.
- The base code sits in the footer or behind a delayed tag manager trigger. It then fires late, or never for people who leave quickly. Meta’s recommendation is between the opening and closing
<head>tags, which also reduces the chance that other code blocks it. - A consent banner holds the base code back until the visitor accepts. That is a legitimate setup, and it means the badge stays empty for anyone who declines.
- A second domain is in play. Check the actual URL in the address bar against the domain you verified, especially after a redirect.
Before you change anything, search the page source for the Pixel ID. If it is there, the browser blocked it. If it is not, the template never had it.
When it shows two Pixels or two events
Duplicate code is the first entry in most lists of setup errors, including Meta’s, and the extension makes it visible.
| What the badge shows | What it means | The fix |
|---|---|---|
| Two Pixels loaded | the code exists twice, for example a plugin plus a hard-coded snippet | keep one copy |
| An old Pixel next to yours | a previous agency’s Pixel is still initialised, and every fbq('track') call fires for each Pixel initialised before it | remove it, or send events to one Pixel only |
| An Auto-detected event beside your coded Lead | automatic event detection is on and guessing at the same action | switch off the automatic tracking setting in Events Manager |
| Two Leads for one submission | the Lead is attached to the button click, or the thank-you page fires it on every load | fire after the server accepts the lead, with an ID tied to the lead |
Automatic event detection deserves a note. It is a convenience, and since August 2026 new Pixels set up through Data Advisor arrive with it pre-selected. Beside a coded Lead it can produce a second, unmatchable event, which is why the badge is worth a look after any change to how the Pixel was installed.
Check the Lead, not only the PageView
A PageView proves the base code runs. It says nothing about the event your campaigns optimize for.
- Fire the Lead from the success path of your submit request, not from the button. Validation errors, double taps and a server that rejects the submission as spam all produce a Lead with no lead behind it.
- Pass an event ID with the browser copy, and the same ID from the server, so the two are paired instead of counted twice.
- Test the consent paths separately: decline, accept on arrival, and accept just before submitting. Which events go out differs between them, and a lost Lead is easier to explain before launch.
- Watch what the extension reports for someone who declines. Your CRM still gets the lead and Meta’s browser count does not, which is a consent gap and not a bug. Landing page views have the same property, as landing page views vs link clicks explains.
What the browser cannot tell you
Conversions API events never pass through the browser, so no extension will show them. The server half is checked in Events Manager: the deduplication tab shows whether the two copies pair, and Test Events shows the payload with the identifiers Meta matched.
That is also the moment to look at the whole landing page rather than the tag. A page that fires everything correctly and says something the ad did not promise still produces expensive leads, and AI landing page generators covers what breaks when a page is generated and never checked.
DropCPL serves the landing page and sets the click cookie in the same response, so the tracking that matters is wired before the page renders, and each lead is tied to the ad, the creative and the landing page. The six steps of a full installation are in Meta Pixel setup, the platform-specific versions are in installing the Pixel on WordPress and installing the Pixel on Wix, and the rest of the tracking questions are in Meta Pixel, Conversions API and attribution.