A Meta custom audience that does not work is usually one of five things: it was created yesterday and Meta does not backfill, the Pixel is not sending the event the rule depends on, it was added as a suggestion rather than a hard limit, it is too small to affect delivery, or it carries a restriction that keeps it out of new campaigns. Each has a different fix, and the order below finds it fastest.
Diagnose in this order
- Open Audiences in Ads Manager and read the status. Populating means it is still filling, Ready means Meta has an audience.
- Check whether the rule depends on events or URLs that actually arrive. A rule on a thank-you URL is worth nothing if the Pixel never fires there.
- Look at how the audience is attached to the ad set. Under Add suggestions it is not a limit.
- Compare the audience size with the daily budget. A small list cannot absorb much spend.
- Read the audience’s own page for a restriction notice, and the data source behind it for a category.
Only the first step is about the audience. The other four are about the connection between the audience and the delivery.
Website audiences fill forward, never back
A website audience needs an active Meta Pixel, rules on events or URLs, and a retention period in days. It starts filling the day you create it and Meta does not backfill, so the first day is empty by design and the first week is thin. Meta also warns that it may miss events from some iOS 14 users, which matters more on a small site than on a large one.
Two failures follow from the rules. A rule on a URL path that the dataset never sees will collect nobody, which happens more often than it should when the data source carries a restricted category: core setup strips custom parameters and everything in a URL after the domain, so a rule such as visited /thank-you can populate slowly or stop updating entirely. And an audience built on a conversion event only grows as fast as that event fires, so a form that sends a Lead once a week produces an audience of dozens, not thousands.
An audience builds from the day you create it. If you need the past, upload a customer list instead of waiting for a website audience to catch up.
A suggestion is not a limit
This is the cause people miss, because the ad set looks correct. In a Leads campaign, an audience added under Add suggestions is what Meta calls an Advantage+ custom audience: a signal about who is more likely to respond, which the delivery system may go beyond. Meta itself does not recommend that setup for retargeting.
To make the audience a hard control, choose Further limit the reach of your ads, then uncheck Use as a suggestion. Exclusions behave differently: an excluded custom audience is always a hard limit, which is why the most reliable audience in a lead gen account is often the one that keeps past leads out of prospecting. The full split between inclusions and suggestions is in Advantage+ audience, and turning it off covers the ad set screen.
Customer lists age, match and stop
A customer list is matched against profiles after Meta hashes the identifiers, and the quality of that match decides how much of the list is usable. Audience Manager shows a match score out of ten, and more identifiers mean a better rate: an email alone matches less than an email plus a phone number.
Three failures are common:
- The list is Outdated. Meta marks it that way after 30 days without an update, and a stale list stops reaching people who changed details.
- The match rate is poor because the file carries placeholders, the same dummy email in hundreds of rows, or badly formatted phone numbers.
- The audience is small because the customer file itself is small, which no setting fixes.
Re-upload on a schedule you can keep, monthly is usually enough, and clean the file before you do.
Restrictions that stop an audience cold
Meta flags custom and lookalike audiences that include information its terms do not allow, and the examples it gives are a specific health condition and a credit score. A flagged audience cannot be used in new campaigns, and live campaigns using one should be edited or paused. The options are to remove the information, build a new audience, or request a review.
The data behind the audience can be restricted too. Meta may assign your data sources to categories such as health and wellness or financial service, and some categories bring core setup. That does not delete the audience; it stops the rules that depend on URL detail from filling. Custom audience lists the categories and where they live in the dataset settings.
When the audience works but the results do not
An audience can be perfect and the campaign still disappoint, because a custom audience changes who sees the ad, not what the ad says. Two checks separate the two problems:
- Compare cost per lead on the same creative before and after the audience was added. If nothing moved, the audience is not reaching delivery at all, which points back to the suggestion setting.
- Count the leads that came from people in the audience, in your own system, not in Meta’s. Retargeting often produces cheap leads that never answer the phone.
Checklist
- Is the status Ready, or still Populating on a young audience?
- Does the rule use an event or URL the dataset actually receives?
- Is the audience attached as a limit, not as a suggestion?
- Is the size plausible against the budget?
- Was a customer list updated in the last 30 days, with a decent match score?
- Is there a restriction notice on the audience or its data source?
If the audience passes all six and delivery still ignores it, the problem is upstream of audiences, and campaign setup, budget and delivery covers what else can hold spend back. The next step from a healthy custom audience is usually a lookalike audience, which needs volume to be worth building. DropCPL keeps the list of past leads on your own side, built from first-party form submissions with the ad attached, so the exclusion list does not depend on a platform export going stale.