Tired creative: the signs of fatigue and how to build a testing cycle
Every creative has a shelf life. The ad that brought the quarter’s best results will, sooner or later, start losing steam: the audience has seen it, already decided or simply stopped paying attention. The challenge isn’t avoiding fatigue — it’s spotting it early and having the next creative ready.
How fatigue shows up in the numbers
Look at the metrics together, for the same audience and the same comparison period:
- Rising frequency: the same people are seeing the ad more often.
- Falling click-through rate (CTR): the ad grabs less attention than before.
- Rising cost per result with no meaningful change in budget, offer or competition.
One metric on its own may be seasonality or a tighter auction. All three together are almost always fatigue.
What’s myth and what’s fact about creatives?
Choose your answer for each statement.
Answer to see your score.
Swapping creatives every week is always best practice.
Frequent changes can reset ad sets’ learning phase and stop the algorithm from gathering enough data. Swaps should follow signs of fatigue and a testing calendar, not a habit.
High frequency always signals a problem.
In remarketing campaigns, for small, warm audiences, higher frequency is expected. The warning sign is frequency rising together with falling results.
The same creative can perform very differently across Stories, Reels and feed.
Each placement has its own format, pace of consumption and context. A vertical video designed for Reels usually does better there than a crop of a horizontal piece.
Video always performs better than a static image.
It depends on the product, offer and audience. In some accounts, an image with a clear offer beats video. Only testing tells.
Testing one variable at a time makes it easier to understand what worked.
If the hook, image and offer all change at once, there’s no way to know which change drove the result. One test per variable produces reusable learning.
The testing cycle we use
Write the hypothesis
One sentence with cause and expected effect: “a hook showing the problem in the first 3 seconds will lift click-through rate compared with the product hook”.
Pick one variable
Hook, format, offer, social proof or call to action. Everything else stays the same as the control.
Set a minimum budget and duration
The test needs volume to exit learning. Meta, for example, says in its help center that an ad set usually exits the learning phase after about 50 optimization events in a week.
Set the decision date
Before launching the test, agree when it will be evaluated — and avoid deciding earlier.
Record the learning
Winner, loser and, above all, why. That record becomes the brief for the next creative.
Have the next creative ready before you need it
The cheapest way to deal with fatigue is not to be caught off guard. Keep a bank of approved variations — hooks, formats and offers — and a production pace that matches the budget: the bigger the investment, the faster the audience saturates and the more creatives the account consumes.
How many creatives your account needs per month
The answer depends on budget and audience size. The larger the investment relative to the available audience, the faster people see the same ad several times and the more creatives the account consumes. A practical way to plan:
- track the weekly frequency of prospecting ad sets and note the point at which cost per result starts to rise;
- note how many days, on average, a winning creative takes to reach that point;
- plan production to have replacements ready before then.
Accounts with a high budget and a small audience may need new variations every week; smaller accounts with broad audiences can sustain the same creative for longer.
What to test first
| Variable | Why test it | Example |
|---|---|---|
| Hook (first seconds) | It decides whether people stop scrolling | Opening with the problem instead of the product |
| Offer | Changes the motivation to act | Free shipping vs discount vs gift |
| Format | Each placement favors a format | Vertical video vs carousel vs image |
| Social proof | Reduces distrust among people who don’t know the brand | Customer testimonial vs number of customers |
| Call to action | Adjusts the commitment you ask for | “See styles” vs “Buy now” |
Start with the highest-impact variables — hook and offer — and only then refine format and call to action. Testing button color before validating the offer is spending budget on the wrong question.
How to record what you learn
A test’s result only becomes a company asset when it’s recorded. A simple log, in a spreadsheet or your project manager, should include: the hypothesis, the variable tested, the period, the budget, each variation’s result and the decision taken. After a few months, that history reveals patterns — which kind of hook works with each audience, which offer sells best in each season — and starts to guide the brief for new creatives.
At Manáry, creatives go through your approval before going live, and each cycle’s learning shows up in the follow-up report for paid media.
Frequently asked questions
What is creative fatigue?
It’s the drop in an ad’s performance after the audience has seen it many times. It shows up as rising frequency, falling click-through rate and increasing cost per result.
How often should I change my ads?
When the signs of fatigue appear, not on a fixed calendar. Very frequent swaps can reset campaign learning before they gather data.
What is the ideal ad frequency?
There’s no universal number. In prospecting, watch for the frequency at which cost per result starts to rise; in remarketing, higher frequencies are normal.
Is it worth using AI-generated creatives?
They can speed up producing variations, as long as they respect the brand identity and platform policies and are tested like any other creative.
Have your ads lost steam?
We analyze your account’s frequency, cost and creatives and build a testing calendar with hypotheses and dates.
Want to apply this to your business?
Manáry looks at your real numbers — media, website and CRM — and hands back the priorities in order of impact.
Get the next pieces by email
Calculators, checklists and practical guides. No spam, one-click unsubscribe.