Definition
Frequency is the average number of exposures per person reached. The platform counts impressions, estimates the number of unique people those impressions reached, and divides the first by the second. Both inputs are scoped to a period and a level: frequency for a campaign over a month is a different number from frequency for one ad set over a week, and neither can be derived from the other. The platform's definition of a person follows its identity system, so on a logged-in platform a person is an account, while on the open web it may be a browser or a device. A frequency cap is a control that limits exposures per person over a chosen window; it is a setting, not the metric, although the two share a name.
Impressions ÷ Reach(ratio)
How to interpret it
There is no universal right level, which is why the direction of this metric is context. A short awareness burst to a broad audience needs enough exposures for the message to register but no more; a remarketing audience with a direct-response objective tolerates far more because the people in it have shown intent and the audience is small. Frequency also rises mechanically over time: a campaign that has already reached most of its audience can only add impressions by serving to the same people again, so a rising figure is partly a sign of saturation and partly a sign of a small audience relative to the budget. The useful comparison is against the same campaign's earlier periods and against the signs of fatigue, which the frequency and creative fatigue guide describes: falling CTRShare of impressions that result in a click., rising CPMCost per 1,000 impressions. and falling results per impression together.
The average hides a distribution. Most platforms offer a frequency distribution report that shows what share of reach fell into each exposure band, and that report is where the problem usually shows: a comfortable average alongside a long tail of people exposed many times. Across platforms the figure cannot be added up at all, because each platform counts only its own reach, and the same person counted on two platforms is two people; the cross-platform reach and frequency guide covers the methods for estimating a blended figure.
What moves it
- Audience size against budget. The smaller the audience and the larger the spend, the faster frequency climbs; it is the main driver.
- Length of the period. Frequency over a longer window is always higher than over a shorter one for the same campaign.
- Placements. Automatic placements spread impressions across more inventory and more people, lowering frequency; narrow placements concentrate it.
- Audience overlap. Ad sets or campaigns that target overlapping people each report their own frequency, while the person experiences the sum.
- Frequency caps and optimisation goals. A cap limits the figure directly; a reach objective spreads impressions, while conversion objectives concentrate them on responsive people.
- Creative count. Several distinct creatives in rotation raise the frequency a person tolerates before fatigue, without changing the number itself.
Reading it in context
| Pattern | Likely reading | Next check |
|---|---|---|
| Frequency rising, ReachUnique people who saw an ad. flat | Audience saturated; spend is buying repeat exposures | Audience size, exclusions, budget level |
| Frequency rising, CTR falling, CPM rising | Creative fatigue or saturation | Frequency distribution, creative age, refresh |
| Low frequency, weak results on an awareness objective | Too few exposures for the message to register | Period length, placement spread, audience breadth |
| Several ad sets each at moderate frequency | Overlap may make the person's real frequency much higher | Audience overlap tool, consolidation |
| Different frequency per platform | Each platform counts its own reach only | Cross-platform dedup method, planning cap |
Same spend, two very different frequencies
A campaign delivers 2,400,000 impressions in a month. Ad set A targets a broad audience and reaches 960,000 people, a frequency of 2.5. Ad set B targets a 60,000-person remarketing list and delivers 420,000 of those impressions, a frequency of 7.0. Within ad set B the distribution report shows 18% of people exposed more than 12 times, and that band's CTR is 0.4% against 1.3% for people exposed four times or fewer. The team adds a cap of 3 exposures per 7 days to ad set B, which lowers its monthly frequency to 4.2 and lifts CTR to 1.1% while shifting the saved impressions to ad set A. The reach and frequency calculator reproduces the arithmetic for any impressions and reach pair.
Related metrics
ReachUnique people who saw an ad. is the denominator and the metric frequency trades against: for a fixed number of impressions, more reach means lower frequency, and the balance between the two is a planning decision that the media planning framework sets before a campaign starts. CPMCost per 1,000 impressions. tends to rise with frequency on auction platforms, because repeat impressions to a saturated audience are bought in more competitive auctions or against people the platform's model rates as less responsive. CTRShare of impressions that result in a click. is the earliest symptom of fatigue, falling as the same people see the same creative again. Impressions are the numerator and, on their own, say nothing about how many people were behind them; effective frequency and effective reach, the planning versions of these metrics, describe the share of people exposed at least a chosen number of times.
For small audiences such as past visitors, the reach and frequency calculator for retargeting shows how quickly frequency climbs and how to choose a cap from the curve.
Common mistakes
- Applying one target frequency to every campaign. Objective, format, audience temperature and period all change what is appropriate.
- Reading only the average. The distribution report shows the over-exposed tail that the average hides.
- Summing frequency across ad sets or platforms. Overlapping audiences make the real figure higher than any single report shows.
- Comparing frequency across different period lengths. A week and a month are not comparable.
- Fixing fatigue with a cap alone. A cap limits exposures; fresh creative and a larger audience address why the exposures stopped working.
- Setting a cap at the wrong level. A cap on each ad set does nothing about the person who sits in several of them; cap where the audiences overlap, usually the campaign or the account.
- Reading a low figure as healthy by default. On an awareness objective, too few exposures is the more common failure, and it is invisible unless effective reach is reported.
