Frequency is creeping. CPMs drift up on the same creative. And the refresh decision rests on a gut feel and a frequency rule someone tweeted in 2022. Sentrum models fatigue as a curve from your own account’s history, grades every creative with the evidence shown, and tells you the refresh cadence your account actually needs.
Not a frequency alarm — a model of how creative actually wears out in your account, graded with the evidence attached.
Sentrum fits decay curves to your own account’s history, per concept and format — how long UGC holds versus statics, where the knee lands, how the CPM-to-result trend bends as an audience wears out. Your account’s curve, not a generic threshold.
Fatigue risk, hook strength, efficiency, consistency — composed into a grade you can scan in a second and expand into the metric, value, and threshold behind every sub-score. No black-box percentiles; the evidence trail is the product.
Decay speed gets compared against your real refresh cadence. If concepts break at day 16 and refreshes land at day 31, Sentrum quantifies what those 15 days of decayed delivery cost — and what cadence closes the gap.
Each signal is computed from your own account’s baselines and gated by significance — so a refresh recommendation means the decay is real.
The cleanest fatigue signal: the same creative buying the same result for steadily more. Measured against the creative’s own history, gated so one bad day doesn’t trigger a refresh.
Concepts in your account have a measurable lifespan by format. Knowing an ad is at day 14 of a 16-day curve turns “should I refresh?” into a scheduling question.
A fading winner and a never-was look identical in a spend report. Hook and hold rates split them — one earns a refresh brief, the other earns retirement.
The account-level read: how fast creative decays versus how fast you actually replace it. The gap between those two rates is quantifiable wasted spend — and the strongest argument for a cadence change you’ll ever put in a deck.
The detection framework is public. Run it by hand on your next suspect creative — or let the scorecards run it on all of them, daily.
The manual detection guide behind this page — the signals, the thresholds, and the refresh playbook.
READ THE GUIDEWhere fatigue fits among the six drivers of any performance drop — and how to rule the others out.
READ THE GUIDESingle-metric rules (“refresh at frequency 3”) can’t tell the difference — every account wears creative out at its own rate. Sentrum models each concept’s decay as a curve from your own account’s history and flags fatigue when the trend is statistically real: rising CPM-to-result alongside position on the curve, not a one-day CTR wobble.
Survival-curve position per concept (how long this format historically holds in your account), the CPM-to-result fatigue trend, efficiency versus the creative’s own baseline, and hook/hold rates to separate wear-out from a creative that never worked. Each signal feeds an evidence-graded scorecard you can expand and inspect.
On your account’s cadence, not a calendar habit. Sentrum compares how fast your concepts decay against how fast you actually ship refreshes — if concepts break at day 16 and refreshes land at day 31, that gap is quantifiable wasted spend, and the recommended cadence comes with the math attached.
No. Meta’s flag fires after delivery has already degraded, and it doesn’t say what to do beyond “try new creative.” Sentrum’s read is ahead of the cliff — position on the decay curve, the cost of the decay so far, and which concept styles in your account hold longest, so the refresh brief writes itself from evidence.
No. Read-only by architecture, never touches budgets or delivery. You get the grade, the evidence trail, and the recommended cadence; the pause button stays yours.