Facebook Audience Overlap Calculator

1. Open Audiences in your ad account. Tick the audiences to compare, then choose Actions and Show audience overlap.
2. Enter the size of the selected audience and of one comparison audience.
3. Enter the Overlap number from the tool. Type the percentage instead when that is the only figure you noted.

Facebook audience overlap is the set of accounts that belong to two ad audiences at once. A person who follows your Page and also visited your website sits in both a follower audience and a website audience.

Meta explains why this matters in its Business Help Centre. Ad sets from one advertiser that target similar audiences can end up in the same auction. Meta then enters the ad set with the best performance history and keeps the others out. The help page says overlap is not necessarily a bad thing, but it can lead to poor delivery.

The Audiences tool reports two columns: Overlap, a count of accounts in both audiences, and % overlap of selected audience. You can compare up to 5 audiences at a time. This calculator works on one pair and adds the figures the tool leaves out.

It is not a delivery forecast. It tells you how large the shared group is, from both sides.

The calculator applies the inclusion-exclusion rule for two sets to your three numbers.

1. It divides the overlap by the selected audience. That is the percentage Meta shows.
2. It divides the same overlap by the comparison audience. The count stays the same and the percentage changes.
3. It adds both sizes and subtracts the overlap once, so nobody is counted twice.
4. It subtracts the overlap from each size to give the accounts unique to each audience.

The Jaccard figure is one number for the pair. It does not depend on which audience you call selected.

Example values only. A selected audience holds 500,000 accounts. A comparison audience holds 300,000. The tool reports an Overlap of 80,000.

The two audiences add up to 800,000 on paper, but only 720,000 different accounts exist. The shared group is a larger part of the smaller audience: 26.7% against 16.0%.

Which Audience Is Selected

The percentage uses the selected audience as its base. Meta gives an example with 1,000,000 and 100,000,000 accounts and 800,000 in both. The overlap is 80% of the small audience and 0.8% of the large one.

Audience Size

Meta says the tool only gives useful overlap information for audiences of at least 10,000 accounts. Sizes shown in the ad account are estimates, so small differences carry little meaning.

How the Audiences Are Built

Two audiences drawn from the same source share more accounts. A website visitor list and a list of past buyers usually share the buyers. Two interest audiences in unrelated topics share fewer.

Exclusions

Excluding one audience from another ad set removes the shared accounts from that ad set. With 80,000 excluded from the 300,000 audience, 220,000 remain.

Time

Audiences that update, such as 30-day website visitors, change size each day. Take both sizes and the overlap on the same day.

The two measures answer different questions about the same shared group.

In the worked example the percentages are 16.0% and 26.7%, and the Jaccard figure is 0.111. A Jaccard value of 1 means the two audiences are identical. A value above 1 is impossible, because the overlap cannot be larger than the combined audience.

To express any of these as a simple ratio, use the ratio calculator.

Before Launching Parallel Ad Sets

Check each pair of audiences that will run at the same time. A pair at 45% and 60% shares most of the smaller audience, which is the case Meta describes as likely to affect delivery.

Sizing a Combined Campaign

Use the unique combined figure, not the sum. It is the right base for reach planning and for the CPM calculator.

Planning an Exclusion

The audience-only counts show what is left after an exclusion. Check that the remaining audience is still large enough to run.

Reporting to a Client or Team

Copy the report with both percentages. Pair it with spend results from the Facebook ad budget ROI calculator.

1. Reading One Percentage as the Whole Story

An overlap of 0.8% of a large audience is 80% of a small one in the Meta example. Read both sides.

2. Adding Audience Sizes Together

The sum counts shared accounts twice. Subtract the overlap once.

3. Treating a Fixed Percentage as an Official Limit

Meta publishes no safe or critical overlap percentage. Lines such as 20% or 30% are rules of thumb from third parties. Set your own alert level and test it against your results.

4. Turning Overlap into Wasted Budget

Multiplying overlap by daily budget does not give money lost. Meta keeps your ad sets from competing in the same auction, so the effect shows up in delivery and not as a fixed share of spend.

5. Comparing Very Small Audiences

Under 10,000 accounts the tool does not give useful overlap information, in Meta's own words.

6. Entering an Overlap Larger than an Audience

The shared group cannot exceed the smaller audience. The calculator rejects that entry.

The arithmetic is exact for the numbers you enter. Its accuracy depends on the audience sizes and Overlap figure, which the platform estimates.

For three or more audiences, run each pair. Pair results do not add up to a three-way figure.

No official figure exists. Meta says overlap is not necessarily bad and publishes no safe percentage, so judge it against your own delivery results and set your own alert level.

Go to Audiences, tick up to 5 audiences, then click Actions and Show audience overlap. The tool lists the Overlap count and the % overlap of selected audience.

The count stays the same but the base changes. 80,000 shared accounts are 16.0% of a 500,000 audience and 26.7% of a 300,000 audience.

Meta says ad sets from one advertiser competing against each other can drive up costs, so it enters only the one with the best performance history. The stated risk is poor delivery for the others.

Add the two audience sizes and subtract the overlap once. 500,000 plus 300,000 minus 80,000 gives 720,000 unique accounts.

It gives one similarity number for the pair, from 0 to 1. It equals the shared accounts divided by the unique combined audience, such as 80,000 / 720,000 = 0.111.

Meta says useful overlap information needs audiences of at least 10,000 accounts. Smaller audiences return little or nothing.

Exclude one audience from the other ad set, or combine the two audiences into one ad set. The audience-only counts on this page show what remains after an exclusion.

The platform tool lets you tick up to 5 audiences and compares each with the selected one. This calculator handles one pair at a time, so repeat it for each pair.

No. The maths runs in your browser and nothing is sent to us. Save keeps a result only in this browser.

• About overlapping audiences: the tool, both columns and the 10,000 account note (Meta Business Help Centre, read 2026-10-02; opens in a browser, blocks automated checks).
• Laws of inclusion-exclusion for counting two sets (Doerr and Levasseur, Applied Discrete Structures, LibreTexts).
• Jaccard coefficient: matches divided by matches plus mismatches (NIST Dataplot Reference Manual).