A quantitative methodology for analyzing competitor ad archives to identify creative burnout, evergreen winners, rapid iteration trees, and dying offer angles before testing your own budget.
Basis for this note
The examples come from the documented commands, tests, limitations, and operating controls in the relevant source package.
What matters
- Ad longevity (>90 days active) is the single highest-probability proxy for profitable ROAS in direct-response advertising.
- A sudden spike in iteration velocity (30+ micro-variants launched in 48 hours) signals that a primary winning hook is experiencing severe fatigue.
- Copy permutation trees reveal whether an advertiser is testing new core angles or merely cycling cosmetic headline variations.
- Tracking rapid kill rates (<4 days active lifespan) highlights failed hypotheses and prevents your team from repeating competitor mistakes.
Signal 1: The longevity baseline (The 90-day survival filter)
In paid social media buying, unprofitable direct-response ads are killed within days. When an ad remains continuously active for 60, 90, or 180+ days across major markets, it is generating positive contribution margin. This is the cornerstone principle of competitive intelligence.
By querying competitor ad libraries and calculating active_days = (current_date - started_date), growth teams can immediately filter out noise and isolate the top 5% of evergreen winners. These long-running assets reveal the competitor’s core value proposition, primary customer avatar, and most resilient proof mechanisms.
Signal 2: Iteration velocity spikes (The fatigue panic indicator)
When a brand’s primary creative begins fatiguing—manifested as rising CPMs, collapsing CTRs, and degraded conversion rates—media buyers respond by rapidly testing minor variations of the winning concept.
In the Ad Library, this appears as an explosive cluster of 20 to 50 near-identical creatives launched on the same date: identical visual footage paired with subtle headline swaps, altered font colors, or revised first-3-second hooks. Detecting this velocity spike tells you that the core angle is exhausted and the market is primed for a fresh visual counter-position.
Signal 3: Copy permutation trees and angle decay
Analyzing primary text variations across an advertiser’s catalog reveals their angle testing matrix. Sophisticated advertisers map copy across distinct emotional buckets: pain-agitation, social proof/testimonial, founder origin story, and direct comparison/switching cost.
When an advertiser stops publishing pain-agitation copy and pivots entirely to heavy price-discounting or urgency-driven claims, it indicates that broad-market cold acquisition has plateaued, forcing them down-funnel into aggressive retargeting and margin-eroding promotions.
Signal 4: Fast-kill mortality curves (Learning from failed spend)
Equally valuable as finding winners is identifying what failed. By filtering historical inactive ads with an active lifespan of under 5 days, you reconstruct the competitor’s failed creative experiments.
If a competitor launched 10 video iterations featuring a specific aesthetic (e.g. 3D motion graphics or high-gloss studio cinematography) and killed all of them within 72 hours while doubling down on raw user-generated content (UGC), you save thousands in production costs by avoiding the rejected format.
Signal 5: Dynamic creative testing (DCT) fingerprinting
When an ad card displays "This ad has multiple versions" with combinations of 3-5 images/videos and 3-5 text bodies, the advertiser is running Meta Dynamic Creative Testing (DCT) or Flexible Creative modules.
Tracking which headline and visual combinations survive DCT incubation into standalone dark-post scaling ads gives you a direct window into Meta’s algorithmic optimization preferences for that target demographic.
Competitor creative fatigue diagnostic checklist
- Ingest full advertiser ad archive and compute active duration (days) for every ad ID.
- Tag and isolate evergreen winners active for greater than 60 consecutive days.
- Calculate 30-day rolling launch velocity to identify sudden variant clustering.
- Cluster ad copy text by semantic hook category (problem-focused, proof, discount, comparison).
- Isolate short-lived dead ads (<5 days) to map rejected angles and creative formats.
- Generate a synthesis matrix identifying market saturation points and white-space opportunities.
pullmesh package
Meta Ad Library collection console
Review the documented commands, current checks, exclusions, platform risk, and purchase terms for the package discussed here.
design partner pilot · 24-hour turnaround
Free 24-Hour Meta Creative Decision Brief
For growth teams spending $20k+/mo on Meta: receive 500–1,000 live competitor ads, 65 structured fields, and a 3-tab executive decision workbook. Free for the first 3 qualified design partners.
Common questions
Why do competitors keep ads active for 6+ months in the Ad Library?
Ads running continuously for 6+ months have accumulated massive social proof and continue to clear performance target ROAS thresholds. They represent the advertiser’s most profitable core creative assets.
How can you distinguish between a new winning ad and a fatigued test variant?
New winning ads typically receive isolated scaling spend and remain active across multiple monthly review cycles, whereas fatigued test variants are launched in high-volume batches and deactivated within 7 to 14 days.
Can you use creative fatigue analysis to inform your own ad briefs?
Yes. By identifying what angles are currently saturated and fatiguing across the category, you can craft counter-positioning briefs that stand out visually and conceptually in the user feed.