Meta has added a new filter to engagement-based custom audiences that gives advertisers more control over who qualifies for a given remarketing segment. It’s called the Engagement Frequency Filter, and it introduces two new ways to define engagement custom audiences for Facebook Pages and Instagram professional accounts.
This post covers exactly what the feature does, which audience types it applies to, the specific scenarios where it adds genuine value, and the broader context that determines whether it’s worth building into your campaign strategy.
What Is the Meta Engagement Frequency Filter?
When creating an engagement custom audience in Meta Ads Manager, advertisers now see two new filtering options:
‘In the past’ filter
This allows you to specify a precise time window for engagement. For example: people who engaged with your Facebook Page in the past 14 to 21 days. This is more precise than the standard fixed-window options (past 30 days, past 60 days, etc.) and allows you to isolate recency more exactly.
‘At least’ filter
This allows you to set a minimum engagement frequency. For example: people who engaged with your Instagram professional account at least 3 times in the past 90 days. This isolates depth of engagement — people who have interacted with your content multiple times — rather than just anyone who engaged once within a time window.
Which Audience Types Does It Apply To?
The Engagement Frequency Filter applies to the following engagement custom audience types:
Facebook Page audiences
- People who engaged with your page
- People who visited your page
- People who engaged with any post or ad
- People who clicked any call-to-action button
- People who sent a message to your page
- People who saved any post
Instagram professional account audiences
- People who engaged with this professional account
- People who visited this professional account’s profile
- People who engaged with any post or ad
- People who sent a message to this professional account
- People who saved any post or ad
That covers the full range of on-platform engagement signals for both Facebook and Instagram, giving advertisers the ability to apply frequency and recency filtering across virtually all standard engagement audience types.
The Context You Need to Understand First
To use this feature effectively — and to avoid over-investing in it where it won’t move the needle — it helps to understand what Meta’s algorithm already does with engagement data automatically.
Since Meta introduced Advantage+ Audience, the algorithm prioritises warm audiences — people who have previously engaged with your brand, visited your website, or interacted with your ads — as part of its standard budget distribution. It does not need to be manually instructed to favour people who have engaged with your page. It already knows, and it already weights budget distribution accordingly.
This is a meaningful shift from how things worked several years ago. Manual engagement audiences were once a core strategy because the algorithm needed explicit signals to prioritise remarketing groups. Today, advertisers who use audience segment breakdowns in Ads Manager typically find Meta allocating 20 to 25 percent of campaign budget to remarketing audiences automatically — without any manual custom audience setup.
The practical implication: for most campaigns, the algorithm’s automatic remarketing prioritisation already covers what manual engagement audiences were designed to achieve. The Engagement Frequency Filter is most valuable in the specific scenarios where that automatic prioritisation isn’t sufficient.
When the Engagement Frequency Filter Adds Real Value
High-intent, high-ticket offers
For products or services with a significant price point or a long consideration period, the quality of your remarketing audience matters as much as the size. Building an audience of people who have engaged at least three times in the past 60 days — rather than anyone who clicked once six months ago — creates a tighter, more committed group that is more likely to be in an active consideration phase. A high-frequency filter applied to message senders or CTA clickers can produce a particularly high-intent segment for direct-response offers.
Specific action-based retargeting
If you want to retarget people who took a specific type of action — sent a message, saved a post, clicked a CTA button — a manually defined engagement audience remains the only way to isolate that group precisely. The algorithm’s automatic remarketing is based on aggregate signals; it doesn’t segment by engagement type. Manual audience construction is still the right tool for this level of granularity, and the frequency filter makes those audiences more precisely defined.
New accounts or limited conversion data
For accounts that haven’t yet generated enough conversion history for the algorithm to prioritise remarketing audiences reliably, manually defined engagement audiences provide an explicit starting signal. Using a frequency filter to ensure the audience reflects meaningful engagement — rather than a single accidental interaction — improves the quality of that signal.
Specific re-engagement campaigns
If you want to run a time-sensitive offer or re-engagement message to people who were recently active with your content — within the past two to three weeks — the ‘in the past’ filter gives you that precision. This is useful for promotional campaigns where recency of engagement is a proxy for current purchase intent.
When It’s Unlikely to Change Much
For established campaigns running with sufficient conversion data and optimising for meaningful goals like purchases or qualified leads, the new filter is unlikely to produce a significant change in results. The algorithm is already prioritising the right engagement signals automatically. A manually defined version of the same audience will largely overlap with what Meta is already doing — and the algorithm’s version typically uses a richer set of signals.
Building complex engagement audience stacks — multiple ad sets each targeting different engagement frequency segments — recreates the kind of campaign complexity that Meta’s current approach is specifically designed to make unnecessary. The audience segments will likely overlap significantly, and the additional ad sets will compete with each other rather than reaching genuinely distinct groups.
How to Use It Effectively
- Use the ‘at least’ filter for high-intent remarketing — people who’ve engaged 3+ times are meaningfully more committed than single-interaction engagers
- Use the ‘in the past’ filter for time-sensitive or promotional campaigns where recent engagement signals current interest
- Combine both for tightest intent signals — e.g. people who sent a message at least once in the past 30 days
- Apply to message senders and CTA clickers first — these are the highest-intent engagement signals available
- Use these audiences as an addition to broader campaigns, not a replacement for them — they are supplemental precision tools, not a primary strategy
- Verify against audience segment breakdowns in Ads Manager before building — if the algorithm is already allocating 25% of your budget to warm audiences, assess whether a manual audience adds genuinely new reach or just duplicates what’s already happening
Key Takeaways
- Meta’s Engagement Frequency Filter adds two new options to engagement custom audiences: ‘in the past’ (precise time window) and ‘at least’ (minimum engagement count)
- It applies across all standard engagement types for both Facebook Pages and Instagram professional accounts
- Meta’s algorithm already prioritises warm engagement audiences automatically for most campaigns — verify this with audience segment breakdowns before building manual audiences
- The filter adds most value for high-intent remarketing, action-specific retargeting, new accounts, and time-sensitive re-engagement scenarios
- For most established campaigns, it is a precision tool for specific situations — not a reason to rebuild a general remarketing strategy around manual engagement segmentation
FAQs
1. Is the Engagement Frequency Filter available for all Meta custom audience types?
It applies specifically to engagement-based custom audiences for Facebook Pages and Instagram professional accounts — covering the standard range of engagement types including page visits, post engagement, CTA clicks, messages, and saved posts. It does not apply to website custom audiences, customer list audiences, or app activity audiences.
2. Do I still need to manually create engagement audiences if the algorithm already prioritises remarketing?
For most established campaigns, no — Meta’s algorithm automatically prioritises warm audiences and allocates roughly 20 to 25 percent of budget to remarketing groups without manual setup. Manual engagement audiences add value in specific scenarios: high-ticket offers requiring tight intent signals, action-specific retargeting, new accounts with limited conversion data, and time-sensitive re-engagement campaigns.
3. What’s the most useful combination of these filters?
For high-intent remarketing, combining ‘at least’ with a meaningful engagement type produces the strongest signal. For example: people who sent a message to your Instagram professional account at least once in the past 30 days. This combines recency and an intent-heavy action type (messaging is a high-commitment engagement signal) into a tight, high-quality segment.
4. Will building engagement frequency audiences improve my ad performance?
It depends on your campaign setup. If Meta is already effectively prioritising your warm audiences through automatic optimisation, a manually defined version of the same group is unlikely to improve performance significantly. If you’re targeting a specific high-intent segment for a specific offer, the frequency filter can help you build a more precise audience than was previously possible.