- Set the reporting basis. Use one consistent campaign or reporting period.
- Enter the source figures. Copy the requested counts from the same analytics view.
- Check units. Enter percentages as whole percentage values when a percent field appears.
- Review the live result. Compare the headline metric with both supporting figures.
- Reset for another scenario. Use Reset to restore the sample values.
Facebook Frequency Calculator
Find the average Facebook ad frequency by dividing impressions by unique reach. The result estimates how many times each reached person saw an ad during the selected reporting window.
Media buyers can compare frequency across ad sets and spot possible overexposure, but the average does not show how impressions are distributed among individual people.
Facebook inputs
Result
—
Average frequency
Repeat impressions—
Reach / impressions—
Formula
Average frequency = Impressions ÷ Unique reach
All counts must cover the same period. The calculator retains full precision internally and rounds only displayed results.
What the result means
Use the headline metric with the supporting figures to understand both scale and efficiency.
Results depend on the definitions, date range, and attribution settings used in the source data.
Given and calculation
With 250,000 impressions and 100,000 people reached, average frequency is 250,000 ÷ 100,000 = 2.5. Each reached person saw the ad 2.5 times on average.
Result
The displayed values provide a planning or reporting estimate based on the inputs supplied.
What reporting period should I use?
Any period works if every input covers exactly the same dates. Monthly periods are often convenient for trend comparisons.
Can I compare separate campaigns?
Yes. Calculate each campaign separately so different audiences, budgets, and attribution windows are not mixed.
Why can this differ from the platform dashboard?
Dashboards may apply rounding, privacy thresholds, attribution settings, or delayed processing. Match definitions and export times before investigating a gap.
How should zero values be handled?
A genuine zero can be entered for optional counts or costs. A denominator required by the formula must be above zero.
Does the result predict future performance?
Only projection-style outputs extend an observed rate, and that assumption may not persist. Treat the result as a scenario, not a guarantee.