Two years ago, a Meta creative concept could hold its audience for 6 or more weeks before decay set in. Under Meta's Andromeda system, that same concept now burns through its audience in 2-3 weeks. The algorithm reads the creative itself, images, copy, video, as its primary delivery signal. It exhausts the best-matched audience as fast as it finds them. A fatigue review built around a six-week lifespan is now looking at creative that decayed twice over before anyone checked it. This post covers why that gap forces a shift from periodic review to continuous scoring, and what continuous scoring does differently.
Quick Summary
Meta's Andromeda system uses creative as its primary delivery signal, so a stale concept does not just lose efficiency, it loses its targeting mechanism.
Effective creative lifespan has compressed from 6+ weeks to roughly 2-3 weeks under Andromeda.
A weekly or quarterly review cadence built for the old lifespan now checks creative well after its decay window has closed.
Continuous scoring tracks the rate of decay across consecutive days, not a single snapshot. That is what catches fatigue while there is still time to act.
Continuous scoring only helps when there is enough conversion volume per creative to make a daily signal meaningful.
Why Creative Now Fatigues in Weeks, Not Months
Andromeda changed what determines who sees an ad. The system reads a creative's images, copy, and video to decide targeting and delivery, not just the audience settings around it.
That mechanism finds a concept's best-matched users fast. It burns through them just as fast. A concept that held an audience for 6 or more weeks before Andromeda now exhausts it in roughly 2-3 weeks. When that happens, the algorithm does not just serve the ad less efficiently. It loses the signal it was using to find the right people at all.
Your Review Cadence Is Already Behind
A fatigue check built for a six-week decay window made sense when creative actually decayed on that timeline. Under a 2-3-week burn cycle, a weekly review falls behind by half of the creative's effective life. A quarterly re2-3-weeks behind by several full cycles.
The practical effect is a report that always looks worse than expected. Engagement on a fatigued ad drops 20-30% per week once decline begins. A creative reviewed even seven days late has already absorbed a real chunk of that decline. The review cadence, not the creative, is usually the thing that needs to change first.
What Continuous Scoring Actually Does Differently
A continuous scoring system receives daily performance data for every live creative, not a weekly or monthly rollup. It computes a decay trend across consecutive days for each one, rather than a single current-state snapshot.
That trend gets compared against a threshold tuned to the account's typical decay curve. Concepts crossing the threshold get flagged automatically. Concepts holding steady or improving surface as candidates for more budget. The output is a ranked, daily-updated list of what to scale and what to refresh, not a static report someone reads a week after the data was current.
A Snapshot Score Misses Fatigue Before It Costs You
Most teams assume a fatigue alert works the same way no matter how often it runs. That assumption misses the actual shape of the decay curve under a 2-3 week burn cycle.
Fatigue is a continuous depletion process. It is not a discrete event a single check will reliably catch at the right moment. A snapshot taken on the wrong day can miss the early inflection point entirely, catching the drop only after engagement has already fallen for a week or more. Tracking the rate of change across several consecutive days, not one data point, is what surfaces the inflection point while a refresh can still happen.
When Continuous Scoring Does Not Help
Three conditions limit what a daily scoring cadence can tell you. An account running fewer than roughly 30-50 conversions per creative per week produces a daily signal too thin to separate real decay from ordinary noise.
A small number of genuinely evergreen brand assets will not show the same 2-3 week decay pattern as acquisition creative. They do not need daily review. Scoring creative faster than a team's production pipeline can supply new concepts just produces an accurate, timely flag with nothing ready to replace the fatiguing asset.
Systems like Maino separate decision logic from execution. They surface which creative needs a refresh without waiting for a scheduled review to catch up. Maino.ai has reduced Customer Acquisition Cost (CAC) by an average of 46% across its client portfolio, a result that depends on catching decay while a swap-in is still possible.
Frequently Asked Questions
What is Meta Andromeda and how does it affect creative fatigue?
Andromeda is Meta's delivery system that reads a creative's images, copy, and video as its primary targeting signal. It finds a concept's best-matched audience fast and exhausts that audience fast, compressing effective creative lifespan from 6+ weeks to roughly 2-3 weeks.
How often should you review creative performance now?
Daily scoring, not a weekly or quarterly review, matches the actual decay window under Andromeda. A slower cadence checks data that has already decayed by the time anyone looks.
What's the difference between continuous scoring and a fatigue alert?
A single fatigue alert reacts to one snapshot crossing a threshold. Continuous scoring tracks the rate of decay across consecutive days for every live creative, catching the inflection point earlier than a one-time check can.
When does continuous scoring not help?
Accounts below roughly 30-50 conversions per creative per week generate a daily signal too noisy to trust. It also adds little value for evergreen brand assets that do not follow the same fast-burn pattern as acquisition creative.
Should every ad account switch to daily creative review?
Only if conversion volume per creative supports a daily signal and the team can produce replacement concepts at the pace flagged. Otherwise, daily scoring just reports the same fatigue sooner without a way to act on it.