What is Ad fatigue?

Ad fatigue is the fall in response that follows repeated exposure of the same people to the same creative. It is not a platform penalty and it is not a mysterious force: it is the ordinary consequence of showing somebody something they have already seen and already decided about.

The people who were going to act on a particular ad mostly act early. What remains, after several exposures, is an audience that has seen the ad and not responded — and each additional exposure is aimed at a population increasingly composed of people who have already declined.

The observable result is a falling click-through rate against rising frequency, followed later by a rising cost per click and, later still, a rising cost per result.

The sequence is consistent enough to be useful as an early-warning system.

Click-through rate falls first. It is the most direct measure of response and it moves within days.

Engagement follows, where the platform reports it, and on social placements negative feedback — hidden ads, reports, unfollows — begins rising at the same time.

Cost per click rises next, because the impression price is being divided by a smaller response rate. Auctions that rank on expected value compound this: an ad the platform now expects to be clicked less wins impressions at a worse price.

Cost per result moves last, often a fortnight after the click-through rate turned.

Accounts monitored on cost per result alone therefore discover fatigue about two weeks after it started, having spent that fortnight paying progressively more for the same outcome. Monitoring the click-through and frequency pair recovers that time.

Both fatigue and an auction change raise the cost per click, and the remedies are completely different, so distinguishing them is the first diagnostic step.

The identity does the work: cost per click is the impression price divided by the click-through rate.

If the impression price is broadly stable and the click-through rate is falling, the response changed. That is fatigue, and the fix is creative.

If the impression price is rising and the click-through rate is stable, the market changed — a competitor arrived, a season turned, demand rose. The fix is bidding, budget or targeting, and rotating creative will do nothing.

If both moved, both things happened, which is common in seasonal peaks where competition and exposure rise together.

Frequency causes fatigue, and an average frequency is a poor instrument for detecting it, because exposure is distributed unevenly.

Heavy users of a platform generate far more opportunities to be served an ad than light ones, so a campaign sitting at a comfortable average shows most people one or two and a minority a great many. The fatigue, the negative feedback and the falling response all concentrate in that minority, and the average conceals them.

Where a frequency distribution is available it is the better report. Where it is not, the turning point in the click-through rate is the practical substitute: it is the account observing its own fatigue directly rather than inferring it from a threshold somebody else published.

Fatigue attaches most strongly to what is visually recognisable. The reliable resets are therefore visual.

A different image or video resets the most, because recognition is primarily visual and a new execution reads as a new ad.

A different format — static to video, single image to carousel — resets substantially, and also reaches different inventory.

A different angle on the same proposition resets meaningfully and is usually the highest-value option, because it also tests something.

A new headline on the same image resets least. It buys a short reprieve and the decline resumes, which is why headline-only rotation produces a sawtooth that trends downward.

Refreshing the offer rather than the execution is a separate intervention and usually the wrong first move: the proposition is rarely what people tired of.

Creative rotation addresses the symptom. Three structural changes address the cause.

A larger audience. Frequency is impressions divided by reach, so adding people lowers repetition without reducing delivery. This is the only intervention that does.

Excluding converters. Continuing to advertise to somebody who has already bought is the most avoidable exposure in any account, and it is astonishingly common in prospecting campaigns with no customer-list exclusion.

Shorter retargeting windows. A ninety-day membership window with a steady budget guarantees very high frequency against whoever joined the pool first.

A frequency cap is the fourth option and it is a trade rather than a fix: it limits exposure by declining delivery, which on a saturated audience means unspent budget.

The useful output of watching fatigue is a rotation cadence specific to the account.

Find the point where the click-through rate turns in a typical campaign. That is the account's own fatigue horizon, expressed in days or in frequency. Schedule new creative to arrive before it rather than after, and the account stops experiencing fatigue as an event and starts treating it as a production schedule.

That schedule moves with audience size and spend rate. A campaign whose budget doubles reaches the same frequency in half the time, and its rotation cadence should halve with it.

Everything above applies more sharply to retargeting, because the audiences are small by construction.

A remarketing pool of a few thousand people receiving a steady daily budget will reach double-digit frequency inside a fortnight. There is no targeting change available that makes the pool larger, because the pool is defined by behaviour rather than by interest.

The controls that work there are different: shortening the membership window so the pool refreshes, excluding people who converted, capping the daily spend in proportion to the pool size rather than to the account's ambition, and rotating creative on a much shorter cycle than prospecting uses.

The failure mode is an account that has correctly identified retargeting as its most efficient campaign and funds it accordingly, then watches the efficiency decay as the same few thousand people are shown the same ad forty times. The efficiency was a property of the audience, and spending more against a fixed audience consumes it.

Two things are regularly mistaken for it.

Seasonal decline. Demand falls after a peak and every response metric falls with it. The signature is that impression price, click-through rate and conversion rate all soften together, which fatigue does not do.

Audience exhaustion. The people worth converting in a small audience have converted, and what remains is the residue. This looks like fatigue and new creative will not fix it, because the problem is who is left rather than what they are being shown. The tell is that reach has flattened while spend continues, and the remedy is a larger audience rather than a new ad.

Distinguishing all three before acting saves a great deal of production effort aimed at the wrong problem.

A campaign opens at a 1.9% click-through rate with an average frequency of 1.4 and a cost per click of A$0.62. Four weeks later frequency is 6.8 and the rate is 0.9%. With the impression price unchanged, the click price is the old one multiplied by the ratio of the two rates: 0.62 × (1.9 ÷ 0.9) = A$1.31, more than double, and negative feedback has risen with it.

Spend and conversion volume were roughly flat until the third week, which is when the cost per result finally moved — a fortnight after the click-through rate began falling, and a fortnight after the account could have known.

• Watch click-through rate against frequency weekly. The pair turns before cost per result does, which buys about a fortnight of warning.
• Rotate the execution rather than the offer when response falls. Fatigue attaches to the creative people have already seen, not usually to the proposition.
• Widen the audience before capping frequency. More people lowers repetition without giving up delivery; a cap lowers both.
• Exclude converters from prospecting and retargeting alike. Continuing to advertise to somebody who bought is the most avoidable exposure in any account.
• Track negative feedback where the platform reports it. A rising hide or report rate is a stronger and earlier signal than a falling positive one, and it carries a delivery penalty.