How often should your ads show? The honest answer is a range, not a number, and this article explains how to find yours. It is built around the Ad Frequency and Reach Calculator embedded on this page, which takes your impressions and reach, computes average frequency, and judges it against the sweet-spot ranges for your audience type: cold prospecting 2 to 5 times, warm retargeting 5 to 8 times, and existing customers 1 to 3 times.
The one formula everything hangs on
Five hundred thousand impressions delivered to 125,000 people is a frequency of 4.0: the average reached person saw the campaign four times. Every frequency question is really a question about this ratio's two moving parts: impressions are bought with budget, reach is bounded by audience size, so frequency rises when budget grows into a fixed audience and falls when the audience expands. The calculator also reports a reach rate (reach as a percentage of impressions, 25% in this example) and, if you give it a target reach and a CPM, the impressions and budget required to hit that reach at the middle of your audience's sweet spot: reaching 200,000 cold prospects at the 3.5 midpoint needs about 700,000 impressions, which at a 8,400** budget.
Why the sweet spots differ by audience
Repetition buys memory, but the amount needed depends on the relationship. A cold audience has no context for you: below about two exposures the campaign barely registers, and the calculator flags frequency under the range as under-delivery, wasted in the opposite direction. Warm audiences (site visitors, engagers) already hold the context, so more repetitions are productive, 5 to 8, because the message is a reminder and an objection-handler rather than an introduction. Existing customers need the least, 1 to 3: they know the product, the marginal exposure mostly risks irritation and unsubscribes, and the calculator's retention range is deliberately the tightest. What changes across the three is not the creative's quality but the audience's starting knowledge, which is why a single global frequency target for an account is a category error.
What happens past the sweet spot
Fatigue is not a metaphor; it has a signature the calculator's verdicts are built to anticipate. Within about two points above the sweet spot, it warns: watch CTR and CPM, because click-through decay and rising costs appear in the metrics before the frequency average itself looks alarming (the average also conceals distribution: a reported 5.0 can mean most people saw the ad twice and a minority saw it thirty times). Beyond that, the verdict is blunt, burnout territory: the same people are being charged repeatedly to ignore you, effective CPM inflates as platforms price in negative feedback, and the fixes are structural, not cosmetic: refresh the creative, broaden the audience, cap frequency where the platform allows, or pause the ad set. Rotating genuinely different creative resets much of the fatigue curve; rotating color variants of the same ad does not, because the audience's boredom is with the message, not the palette.
Planning backwards from a reach target
The calculator's planning mode inverts the usual workflow: instead of spending and discovering reach, set the reach the campaign needs (say 200,000 of a defined market), pick the sweet-spot midpoint frequency for the audience type, and derive impressions and budget before launch. This exposes impossible plans immediately: if the target reach times the needed frequency costs more than the budget at realistic CPMs, something must give (narrower geo, longer flight, smaller reach ambition, or accepting the bottom of the sweet spot, 2.0 for cold, which halves the impression requirement versus the top). It also exposes the opposite pathology: a budget so large for the audience that the implied frequency lands in the teens, which is a plan to pay for burnout.
Measurement caveats that qualify every number
Reach and frequency are platform-modeled, not census counts: identity resolution across devices is probabilistic, each platform defines and deduplicates a "person" differently, and cross-platform frequency (the same human seeing you on Meta, Google, and TikTok) is invisible to every individual dashboard, so true total frequency is always higher than any single platform reports. Attribution windows further blur which exposures count. Treat the calculator's verdicts as directional instruments: exact at arithmetic, approximate at reality, and most reliable compared against your own account's history rather than against universal benchmarks.
Frequently Asked Questions
Is a frequency of 4 good?
For cold prospecting, it sits inside the 2 to 5 sweet spot and the calculator calls it balanced. The same 4.0 aimed at existing customers is above their 1 to 3 range and into fatigue-watch territory. The audience, not the number, decides.
My frequency is low but results are poor. Is more frequency the fix?
Not necessarily. Below-range frequency on a cold audience does cap what the campaign can achieve, but poor results at adequate frequency point at creative, offer, or landing page instead. Frequency is a delivery diagnosis, not a performance guarantee.
How often should I check frequency?
Weekly for active prospecting, and whenever budget or audience size changes materially, because frequency moves with both. The trend matters more than any single reading; a fast-climbing frequency on a fixed audience is the earliest fatigue warning the account gives you.