How to Identify a Competitor's Best-Performing Ads (Signals Guide)

The clearest signal a competitor's ad is winning isn't anything they tell you, it's how long it keeps running unchanged. Beyond that, best-performing ads tend to share a small set of observable traits: sustained longevity, a stable core hook reused across creative variants, multi-geo or multi-network placement, and a landing page that stops changing once it's found its groove. None of these are guesses; you can check every one of them directly from what a competitor is publicly running.

Longevity is the strongest single signal#

An ad that's been live for a week could be anything, a new test, a seasonal push, a mistake nobody's caught yet. An ad that's still running after 30 days has survived enough scrutiny (and enough spend) that it's very likely profitable, because almost nobody keeps a losing creative live at cost for a month straight. Our data on ad longevity as a winning signal backs this up, and the longest-running native ads in the OpenAdLibrary index cluster at the high end of observed run times, which is exactly the pattern you'd expect if longevity really does track performance. When you're scanning a competitor's live creative set, sort by run length first. It filters out the noise faster than any other single metric.

Creative variation without angle variation#

A strong ad rarely runs as a single static asset for long. What you'll typically see instead is the same core angle reappearing across several creative executions, a different image, a slightly reworded headline, a different aspect ratio for a different placement, but the same underlying claim and hook. That pattern (angle constancy, execution variation) is a much stronger tell than a single ad's design. It means the advertiser has already confirmed the angle converts and is now optimizing delivery, not still testing whether the idea works at all. Our framework in hook vs angle vs claim breaks down exactly which layer is usually held constant and which layer gets iterated once an ad is performing.

Placement spread across networks and geos#

An advertiser confident in a creative tends to push it wider, not just longer. Watch for the same core offer appearing on Taboola and Outbrain simultaneously, or the same angle showing up translated into two or three languages for different geo targets. Single-network, single-geo presence for weeks on end can still be a winner, especially in a niche vertical, but cross-network and cross-geo spread is one of the fastest ways to confirm an advertiser has moved from testing into scaling.

The landing page stops moving#

Early-stage tests often cycle through several pre-lander variants and offer pages as the advertiser hunts for the version that converts. Once they find it, the landing page tends to stabilize, the same URL and layout persisting alongside a growing set of ad creatives pointing at it. If you trace the redirect chain from several different ad variants and they all resolve to the same stable landing page, that's a strong indicator the funnel underneath the ad is doing real work, not just the headline. Our guide to reverse-engineering a competitor's ad funnel walks through tracing this end to end.

Device and format concentration is a quieter signal#

Look at which device an ad is optimized for, not just whether it exists on mobile and desktop. If a competitor runs a slightly different headline or image crop specifically tuned for mobile, and that mobile-specific version has been live longer than a corresponding desktop version, it usually means mobile is where the offer actually converts for them, and they've stopped bothering to maintain a desktop variant that underperforms. The same logic applies to format: a competitor who's built a dedicated carousel or motion-ad version of an otherwise-static winning creative is investing production time into a format upgrade, which advertisers generally reserve for offers they already trust.

A worked example: scanning one advertiser's live set#

Imagine you pull up a competitor's current creative footprint and see 18 live ads. Twelve of them share one headline structure ("[Symptom]? Doctors Recommend This Instead") with different images, and those twelve have run between 12 and 34 days. The other six are testing a completely different angle and have all been live under a week. Without knowing a single number about spend or conversion rate, you can say with reasonable confidence: the symptom-led angle is the confirmed winner, the advertiser is still in early testing on the second angle, and if you're entering this vertical, the first angle is the one worth studying closely for hook structure and offer framing, not the newer one.

A signal checklist#

Watch for the tell that separates a real winner from a lucky survivor#

Occasionally an ad survives for weeks not because it's converting well but because the advertiser simply forgot to pull it, especially on networks where minimum daily budgets are low enough that an underperforming creative can limp along without anyone noticing. The way to tell the difference is whether the advertiser is still actively investing around that ad: new creative variants being added on the same angle, the landing page getting refreshed, or the offer appearing on additional networks. A 40-day-old ad with no creative siblings, no variant testing, and no expansion is more likely dormant than dominant. A 40-day-old ad that's spawned a dozen variants and expanded to two more networks is very likely a genuine winner.

What this doesn't tell you#

None of these signals reveal the actual conversion rate, cost per click, or margin behind an ad. Two advertisers can both show every signal above and still be running very different economics underneath, one might have a $40 CPA offer with a $60 average order value, the other a $4 lead-gen payout at a much thinner margin. Longevity and scale tell you the ad clears some profitability bar for that specific advertiser; they don't tell you what the bar is or whether the same math would work for you. Treat these signals as "worth studying closely," not "guaranteed to work if copied directly."

Turning signals into a research routine#

The signals above are only useful if you check them on a schedule, not once. A simple weekly routine: pull a competitor's current live creative set, sort by run length, flag anything crossing 30 days for the first time, and note whether the landing page it points to has changed since last week. Over a few weeks this builds a real picture of which of their offers are actually working, versus which ones cycled through and got pulled. Treat it as a standing habit rather than a one-off audit, and see analyzing winning native ad creatives for how to go deeper on scoring the creative itself once you've identified a candidate worth studying.

How OpenAdLibrary surfaces this without manual tracking#

Manually rechecking run length and landing-page stability across a watchlist of competitors every week is exactly the kind of task that's easy to let slide. OpenAdLibrary's native ad spy tool tracks first-seen and last-seen dates automatically across 49 networks, so longevity and creative-variant counts are visible at a glance instead of requiring you to remember what an ad looked like last week. Pair that with ad creative analysis to score the hooks and angles once you've narrowed down which ads are actually worth studying, rather than trying to eyeball every live creative a competitor is running.