"Refresh your creative every two weeks" is the kind of advice that sounds actionable and falls apart the moment you apply it to a real account. A campaign spending $500 a day against a 50,000-person audience wears out on a completely different timeline than one spending $500 a day against a 2 million-person audience. Knowing when to refresh ad creative isn't about counting days on a calendar, it's about reading a small set of signals relative to your specific audience and spend, and building a cadence around those signals instead of a fixed number.
Why there's no universal day-count or spend-count rule
The instinct to want a simple rule (refresh every 14 days, or every $10,000 spent) is understandable, it's easy to plan around. But two campaigns hitting the same day count or the same spend threshold can be in completely different states. A large evergreen audience absorbs the same creative for much longer before frequency climbs to a fatiguing level. A narrow retargeting audience can wear out a creative in days. Any rule that ignores audience size and delivery pace is going to be wrong for a large share of the campaigns it's applied to.
The framework: watch ratios and trends, not calendar dates
Instead of a fixed schedule, use three relative signals to decide when to refresh ad creative for a given campaign.
Frequency relative to audience size
The same frequency number means different things at different audience sizes. A frequency of 4 on a 30,000-person retargeting list is a much stronger fatigue signal than a frequency of 4 on a 3 million-person cold audience. Set your frequency threshold as a function of audience size, not a flat number applied everywhere.
Performance trend, not a single data point
One soft day doesn't mean anything. A rolling multi-day decline in CTR or engagement, especially across a cluster of related creatives, is the signal that actually matters. Judge the trend line, not yesterday's number.
Spend concentration and campaign role
A high-spend, always-on evergreen campaign deserves a proactive refresh cadence built in from the start, because the cost of fatigue compounds daily at scale. A small test campaign or a short flight can usually run its course without a mid-flight refresh.
Set the review cadence, then act on signal, not schedule
Put a recurring review on the calendar (weekly for high-spend evergreen campaigns, biweekly for smaller ones), but let the actual refresh decision be driven by what the frequency, trend, and cluster data show at that check-in, not by the fact that two weeks have passed.
Reactive versus proactive refresh
The framework above naturally splits teams into two camps: those who refresh after performance has already visibly declined, and those who refresh on a scheduled cadence informed by fatigue signals before the decline shows up in the topline numbers. For high-spend, always-on campaigns, the difference between these two approaches is a meaningful amount of wasted spend.
Reactive refresh (wait for decline)
- Refresh is triggered only after CTR or CPA has already visibly worsened
- Days of spend flow into a fatigued creative before anyone notices
- Replacement creative gets built from scratch, under time pressure, after the problem is already showing in reports
- Works fine for low-spend or short-flight campaigns where the downside of a late refresh is small
Proactive refresh (scheduled cadence + fatigue signal)
- A recurring review cadence checks frequency and trend before performance visibly craters
- Replacement creative is drafted ahead of the decline, not scrambled together after
- High-spend evergreen campaigns get refreshed on a rhythm that matches their audience size and delivery pace
- Fits campaigns where even a few days of fatigued delivery represents real budget at stake
Proactive refresh isn't about guessing further in advance, it's about acting on the same frequency-and-trend signals earlier in their arc, before they show up as a visible CTR or CPA problem in the weekly report.
Why proactive refresh usually loses to a bottleneck, not a lack of will
Most media buyers already know proactive refresh is the better approach in theory. It loses out in practice because building a new, on-brand, on-strategy creative brief from scratch takes real time, and that time cost makes "wait until it's clearly declining" the path of least resistance when a team is already stretched across a dozen live campaigns. The framework is right, the bottleneck is production capacity, not judgment.
What makes proactive refresh practical instead of a bottleneck
This is the specific gap Clarifyad's Fatigue → Fix is built to close. When a cluster of creatives crosses a fatigue threshold, Clarifyad drafts a scored, on-brand replacement brief automatically, so there's no blank-page starting point once the signal fires. A one-click approve turns that draft into a ready-to-produce asset, cutting the lag between "we should refresh this" and an actual asset in production from days down to a single decision. That's what turns the proactive column in the comparison above from an aspiration into something a lean team can actually run week over week.
Pairing that with Generation & Testing closes the loop further: creative batch generation recombines historically-winning headline, visual, and CTA components rather than starting the replacement from zero creative direction, and the multivariate testing lab lets you validate the refresh candidate before committing full budget to it. Together, that's the difference between refreshing on a rhythm your account can actually sustain and refreshing only when the numbers force your hand.
Applying the framework to a mixed portfolio
Most accounts aren't running one campaign, they're running a portfolio that mixes always-on evergreen spend with short-flight tests and seasonal pushes. The framework above shouldn't be applied uniformly across that portfolio. High-spend evergreen campaigns earn a proactive, scheduled review because the compounding cost of even a few fatigued days is real money. Short-flight campaigns and small tests often finish before frequency ever climbs high enough to matter, so a mid-flight refresh there is usually wasted effort better spent on the campaigns where fatigue risk is actually concentrated. Sorting your portfolio this way before setting cadences keeps the framework practical instead of turning into a blanket policy applied without judgment.
The takeaway
There's no honest single answer to when to refresh ad creative that applies across every account, audience size and delivery pace change the math too much for a flat rule. What works is a framework: watch frequency relative to audience size, judge trend over single data points, and set a proactive cadence for your highest-spend evergreen campaigns rather than waiting for decline to force the decision. The teams that pull this off consistently are the ones where drafting the replacement isn't the bottleneck.