Notch helps performance marketers and growth teams decode the financial logic of their rivals by treating the Meta Ad Library as a mathematical proxy for spend and customer acquisition costs. To estimate a competitor's budget without internal data, you must analyze three specific metrics: the survival rate of individual creatives, the density of active variants, and the weekly iteration velocity of the account. By calculating the minimum viable learning budget required to sustain their active ad count, you can identify their most profitable angle families and replicate their winning mechanics for your own campaigns in 2026.
Establishing the baseline survival rate at Notch
The single most reliable indicator of a competitor's profitability is the duration an ad remains active. Because growth teams rarely continue paying for losing creative for more than a few days, an ad that survives past the 30-day mark is a statistically probable winner. At Notch, we use this survival rate to filter out the noise of failed experiments and focus on the "anchors" of a competitor's account.
- Ads running for 30+ days: High probability of being at or above break-even ROAS.
- Ads running for 90+ days: Proven scaling winners that likely represent the bulk of their account spend.
- Rapidly disabled ads: Failed hooks or offers that were deemed unprofitable during the initial testing phase.
The technical constraints of the Meta Ad Library mean you cannot see the exact dollar amount behind a creative, but you can see the active dates. When an ad appears on the first of the month and is still delivering on the 30th, the competitor has likely passed the "learning phase" and entered a scaling phase. This is especially true for DTC brands where margins are thin and the tolerance for unprofitable spend is low.
Filtering for the 30-day benchmark
When you audit a competitor, use the "Active Ads" filter and look specifically for ads with the oldest "Started running on" dates. A healthy account typically has a small handful of these long-running winners that provide the baseline performance for the business. These ads have survived multiple algorithm shifts and creative fatigue cycles, meaning their "physics"—the timing, the hooks, and the offer—are highly optimized.
Compare these veterans against the newest launches. If the veterans are cinematic and the new launches are low-fidelity UGC, you are witnessing a pivot in their creative strategy. By tracking the survival rate of these new formats over the next four weeks, you can determine if their pivot was successful or if they returned to their original cinematic baseline.
Interpreting format diversity
Profitability often leaves a footprint in the variety of formats a brand uses. An advertiser running the same concept as a static image, a Reel, and a carousel is usually "multiplying" a winner. They found an angle that converts and are now horizontally scaling it across every possible placement to maximize reach and minimize frequency.
At our San Francisco office, we've analyzed over 5,000 brands and found that format diversity is a lead indicator of production budget. Brands that only run one format usually have a production bottleneck. Brands that run five formats for every one angle have an optimized creative engine that turns a single successful script into a dozen assets.
Measuring testing capacity through iteration velocity in our San Francisco office
The total number of active ads in a library is a snapshot, but the rate at which those ads are replaced—known as iteration velocity—tells you how much the competitor is spending on discovery. According to research on Facebook ad spend tracking, a brand launching 20 or more variations per week is operating with an aggressive testing culture and a significant budget floor.
- Launching 0-5 ads/week: Low testing volume, likely relying on a few static winners or a limited budget.
- Launching 5-20 ads/week: Moderate testing, typical of a healthy Series A/B brand.
- Launching 20+ ads/week: High-velocity testing that requires automated production infrastructure like Notch.
You can reverse-engineer the "testing floor" of a competitor by applying the $3-$5 rule. Performance marketers generally allocate a minimum of $3 to $5 per day per creative to gather enough data for the algorithm to exit the learning phase. If a competitor has 100 active ads, their daily spend floor is at least $300 to $500 just to keep those tests alive.
The math behind minimum viable learning
To build a defensible spend estimate, multiply the number of active ads by the daily minimum viable learning floor. If you see a brand like MyDegree running 200 active variations, you are looking at a minimum daily spend of $600 to $1,000. However, most accounts don't just spend the minimum; they stack budget on the top 10% of performers.
A more accurate calculation involves weighting the spend based on ad duration. Long-running ads (30+ days) likely command 70-80% of the total budget, while the "testing layer" of ads under 7 days old commands the remaining 20-30%. If you identify 10 veteran ads and 90 new tests, you can estimate that the brand is spending heavily to find the next "anchor" that will allow them to scale to the next level of spend.
Weekly volume vs. active density
Total active density can be misleading if a brand is in the middle of a "shotgun" testing phase. If they launched 150 ads yesterday but only have 5 ads from last month, their account is in chaos, not scaling. True velocity is measured by the delta between total ads launched and total ads retained.
High retention of new launches indicates a high "hit rate" on their creative strategy. This usually happens when a brand is extracting creative physics from competitor ads before they ever hit the "launch" button. They aren't guessing; they are testing variations of already-proven hooks, which significantly lowers their effective CAC by reducing the number of failed experiments.
How Notch isolates proven angle families from the noise
Competitors rarely scale a single, isolated ad. Instead, they scale "angle families"—groups of ads that share the same psychological trigger but use different visual executions. When you see a brand running 15 variations of a "Problem/Solution" hook and only 2 variations of a "Unboxing" hook, you have found their primary growth lever.
- Identifying the dominant hook: Look for the recurring first three seconds across the highest-density ad clusters.
- Mapping the visual triggers: Note whether the winners use text overlays, synthetic avatars, or cinematic b-roll.
- Spotting the offer structure: Check if the angle family is tied to a specific bundle, discount, or guarantee.
By identifying these families, you can stop wasting time on broad competitor audits and focus on the 20% of their creative that is driving 80% of their revenue. This systematic approach is how you can effectively find winning competitor hooks using ad library survival rates. You are looking for the "winners of winners"—the specific angles that the competitor is doubling down on with multiple variants.
At Notch, our agents are designed to recognize these patterns autonomously. Instead of a human spending hours scrolling through the library, the system identifies the highest-density clusters and reverse-engineers the script and visual pacing that makes them work. This allows you to skip the expensive discovery phase that your competitor already paid for.
Analyzing the risk envelope
Every performance marketer has a "risk envelope"—the amount of money they are willing to lose on testing before they find a winner. A competitor with a high iteration velocity and a low survival rate has a wide risk envelope, meaning they are well-funded but struggling to find a stable "hit."
Conversely, a competitor with low velocity but high survival rates has a narrow risk envelope. They are surgical in their testing. They likely spend weeks on a single "hero" asset before launching. Understanding which profile your competitor fits helps you decide whether you should try to out-test them on volume or out-think them on creative quality.
Rebuilding the creative physics of long-running winners with Notch
Once you have identified the veteran ads and their corresponding angle families, the final step is to rebuild their creative physics. This isn't about copying the ad; it's about extracting the exact timing, triggers, and hooks that are compounding data for the competitor. If their top-performing ad hits a visual hook at 1.2 seconds and a benefit claim at 4.5 seconds, that "pacing" is a proven asset.
- Extracting the hook: Identify the specific visual or audio "pattern interrupt" used in the first 3 seconds.
- Analyzing the pacing: Measure the "cuts per second" and the timing of text overlays.
- Replicating the "physics": Use an agentic engine to build a net-new video that follows the same structural blueprint but uses your brand's unique assets and voice.
By using Notch Cinematic Shorts, you can take a winning competitor URL and instantly generate dozens of variations that follow the same high-converting structure. This process allows you to map and clone competitor hooks while maintaining your brand's specific identity. You are leveraging their data-backed conclusions without inheriting their creative fatigue.
Extracting hooks and pacing
A winning ad's "physics" is often hidden in the first few frames. In many cases, a competitor is using a "split-screen" hook or a "stop-motion" effect that triggers a higher thumb-stop rate. When you analyze these in the library, look at the "Started running on" date specifically for the variants. If they launched 10 variants of a hook and 8 are still running after 14 days, that hook is a goldmine.
You can then apply these lessons to your own production. If the library shows that their longest-running ads are all 15-second "Cinematic Shorts" rather than 60-second UGC testimonials, you should pivot your testing budget toward shorter, higher-impact formats. This data-driven pivot is far more effective than following general "best practices" that don't account for your specific niche.
Multiplying the proven format
The ultimate scaling move is format multiplication. Once you find a competitor's winning physics, don't just make one ad. Create a testing matrix. At Notch, we recommend taking one proven angle and generating 20 variations: change the background, change the avatar, swap the music, and test five different headlines.
This high-volume approach ensures that even if the competitor's original ad fatigues, your version has enough internal variation to continue performing. You are effectively building an "on-demand ad infrastructure" that turns one competitive insight into hundreds of performance-ready assets.
One thing to watch out for: the recent launch trap
The biggest mistake marketers make in the Meta Ad Library is assuming that a high volume of active ads equals success. If you see 200 active ads but 190 of them were launched in the last 48 hours, you aren't looking at a scaling machine—you are looking at a brand in the middle of a massive (and potentially desperate) testing phase.
Always prioritize the "Active Ads" with the oldest dates. Those are the stable anchors of the account. The "testing layer" at the top is chaotic and unproven. If you clone a test that the competitor is about to kill tomorrow, you are importing their failures into your own account. Wait for the survival rate to prove the concept before you invest your own production budget.
You don't need to see a competitor's bank statement to map their growth strategy. By reading the library with a media buyer's discipline—measuring survival, velocity, and density—you can uncover the exact CAC-lowering mechanics that they are using to scale. Once those mechanics are exposed, use an agentic engine to bridge the production gap and ship your own variations to Meta and TikTok in minutes.