Frequency

Impressions divided by reach. That is the whole definition, and it means frequency is determined by two things you set indirectly: how much money is going out, and how many people are eligible to receive it.

A large budget on a small audience produces high frequency automatically. The platform has to spend the money and there is nobody new to spend it on, so it shows the ad again to the people already in the pool. No campaign setting prevents this, because it is not a setting.

This is why frequency caps so often disappoint. A cap does not create audience; it withholds impressions. The budget then either underdelivers, or the platform reaches further into less relevant inventory to place it, which raises cost and lowers quality. The cap treated the symptom.

The exception worth knowing is that capping is genuinely useful when the audience is large and the goal is efficient coverage rather than saturation. On a small audience it is a way of hiding a budgeting mistake from yourself.

A reported frequency of 3 is compatible with two completely different situations. In one, most people saw the ad three times and the distribution is tight. In the other, the majority saw it once and a small group saw it twelve or fifteen times.

The second is where complaints come from, and it is invisible in the headline. The people writing to the client about being followed around the internet are in the tail, not at the mean.

Most platforms offer a frequency distribution breakdown rather than only the average, and it is worth looking at when a client raises the issue, because the answer is usually that a small subset is being hit hard rather than that everyone is.

The tail also tends to be the most engaged part of the audience, since delivery systems favour people who interact. The ad is being shown most often to the people most likely to be annoyed by it.

Frequency rising while click-through rate falls is the clearest early signal available in a social account. It leads cost per result by a week or two, because the effect has to pass through clicks and conversions before it becomes a cost figure.

The sequence is consistent: frequency climbs, CTR softens, the delivery system responds to weaker engagement by placing the ad less favourably, CPM rises, and only then does cost per conversion move. Catching it at step one or two means a creative refresh. Catching it at step four means explaining a bad month.

This is one of the few genuinely predictive numbers in paid media, and it is routinely missed because frequency is not usually on a monthly report. It belongs on a weekly glance alongside CTR.

Widen the audience first. High frequency almost always means the pool is too small for the money, and adding people is the only change that addresses the cause rather than the symptom. Broader interest targeting, a lookalike, a new geography, or simply removing over-specific exclusions.

Refresh the creative second. New creative resets the fatigue without changing the underlying arithmetic, which buys time and works well when the audience genuinely is the right one and simply needs something new to look at.

Reduce the budget third, and only when neither of the first two is available. It is the option that guarantees fewer results.

For an agency, the first option is the one that needs a conversation, because widening targeting usually means accepting a worse cost per result in exchange for volume. Having that conversation from a frequency chart is considerably easier than having it after two months of decline.