Ad frequency: how many impressions per person actually pay?

Ad frequency is how many times a single person sees your ad in a chosen window, and the math is one line: impressions divided by reach. The profitable band is narrower than most accounts assume. On cold audiences, 2-3 weekly impressions collect most of the return; on remarketing the band opens to 5-8. Past that band, every extra impression buys the same sale at a higher price, and beyond a point the contribution goes to zero and then negative.

What is ad frequency and how is it calculated?

Ad frequency is the average number of times a user saw a campaign inside a defined time window. The formula is simple: Frequency = Impressions / Reach. With 400,000 impressions and 100,000 reach, frequency is 4.0. The detail that matters is the window: 4.0 across a campaign lifetime and 4.0 in the last seven days are entirely different situations. Always read frequency weekly, because published thresholds are defined weekly.

• Impressions: how many times the ad was served. The same person is counted again and again.
• Reach: the number of unique people who saw the ad at least once.
• Frequency: impressions divided by reach, the average repeat per person.
• Effective frequency: the lowest number of repeats that meaningfully lifts conversion probability.

What does the frequency and conversion curve look like?

The curve is not linear, it has diminishing returns. The first impression produces the largest incremental effect, every repeat does less work than the one before it, and at some point the contribution turns negative because budget spent on the same person was taken from reaching a new one. In practice you see three zones: recognition, efficiency and waste. Your only real job is finding where the efficiency zone ends.

• 1-2 impressions (recognition): the message has not landed yet, conversion rate is low but incremental value is highest.
• 3-5 impressions (efficiency): recall overlaps with purchase intent and unit cost is at its best.
• 6+ impressions (waste): click-through falls, cost per result rises, brand fatigue starts.

What is the math between audience size and frequency?

Frequency is not a behaviour, it is a ratio: your budget divided by your audience. Written out: Frequency = (Budget / CPM x 1000) / Reach, where CPM is the cost per thousand impressions. With a fixed audience, doubling the budget doubles impressions, and because reach saturates, the surplus goes straight into frequency. High frequency is usually a scale problem, not a creative problem.

When should you set a frequency cap, and at what number?

A frequency cap fixes the maximum number of impressions one person can receive in a period. Set it just below the point where your marginal contribution curve approaches zero. Three weekly impressions is a reasonable cold-audience starting point, and remarketing can run up to eight. The catch is that this setting does not exist in every campaign type, so where you cannot cap, you have to steer frequency indirectly.

• On Meta: frequency cap and target frequency controls come with awareness objectives and with reach and frequency (reservation) buying. Conversion-focused auction campaigns have no hard ceiling.
• On Google: Display campaigns expose frequency management in campaign settings with a per day, week or month limit, and only viewable impressions count toward it.
• Where no cap exists: Demand Gen and Performance Max have no frequency cap setting, so audience breadth, exclusions, budget pacing and creative count become your levers.
• Rule: cap against your own curve, not against a competitor or a habit. A cap is a budget-steering device, not a safety ceiling.

Why are the thresholds different for cold and remarketing audiences?

The gap comes from intent. Someone in a cold audience does not know you; if the fourth impression has not moved them, a fifth rarely will, and the irritation cost compounds fast. In remarketing the person already visited, viewed a product, maybe added to cart. Intent is live, so reminders keep working for longer, which is why the band roughly doubles. We covered the segment logic in our remarketing strategy guide: giving a cart abandoner the same frequency as a homepage browser is a mistake before the number ever matters.

How do you spot rising frequency early?

The early signal is not frequency itself, it is the diverging speed of impressions and reach. When impressions keep climbing while reach flattens, frequency is mathematically rising, and you can see that weeks before performance breaks. For the performance-side fatigue symptoms, see our piece on early ad fatigue signals; here the focus stays on the number itself.

1. Pull the last seven days of impressions and reach per campaign and divide to get frequency.
2. Repeat the same calculation for the previous seven days and take the difference between the two.
3. Check reach growth as a percentage: if it lags impression growth clearly, frequency is climbing.
4. Plan the intervention before the threshold is crossed, while click-through and cost per result are still intact.

Widen the audience or refresh the creative?

The decision rule is simple: if the problem is in the denominator, change the audience; if interest falls independently of the denominator, change the creative. Widening the audience lowers frequency directly and fast. Refreshing creative does not lower frequency at all, it lets you survive the same frequency for longer. They solve different problems. To structure the creative side, our ad creative testing framework lays out the test design step by step.

• High frequency, flat reach, small audience: widen targeting or add a lookalike audience.
• High frequency, growing reach, falling click-through: refresh the creative and change the message.
• High frequency but ROAS intact: do not touch it. Put it on a watchlist. High frequency alone is not a problem.
• Low frequency but weak performance: frequency is not your issue. Look at bidding, creative and the landing page.

How does Ads Sensor help with the frequency decision?

Ads Sensor unifies Meta Ads, Google Ads, TikTok Ads, Criteo and GA4 data in one panel, so impressions, reach, clicks and conversions sit on the same screen and frequency becomes easy to read per account and per campaign. AI-powered deep campaign analysis reads that table and returns prioritised, reasoned actions; round-the-clock anomaly monitoring catches sudden cost and conversion swings without you watching; and applied recommendations get automatic before and after tracking so you can see whether the call worked. It does not manage frequency automatically on your behalf, the decision stays yours. The beta signup is open if you want to try it.