By Ryan Richardson · Published 8 October 2026
A swipe file of competitor ads gets treated as evidence of what works, sometimes cited to a client as if it were performance data, based on nothing but what's currently live.
Every advertising platform of size publishes what's currently running, and almost nobody checks it properly. What's live today tells you nothing on its own, because launching an ad costs nothing and anyone can do it regardless of whether it's working. Longevity is the only honest inference available: a campaign still running after ninety days is being paid for by something, because nobody keeps funding a loser that long. One $4.99 product in a comparable library ran live for 482 days, which is the kind of signal this exercise can actually give you, and the only kind.
A second version of the same mistake is treating a swipe-file screenshot with no screenshot date and no run duration as evidence, simply because it gets forwarded around enough times to feel true.
Table what you find with the same columns every time: advertiser, days running, who the first line is spoken to, the promise, the price on the destination, the proof type, the awareness stage and the opening words verbatim. Keep one row at the bottom naming what none of it tells you, spend, impressions, conversions, so you never cite any of this to a client as performance data. Read the longest-running advertisement in your category line by line rather than skimming it; what actually does the work is usually the format and the opening move, both reproducible once seen clearly.
If a competitor-ad analysis cites click-through rate, conversion rate, or spend for a competitor's campaign, ask where that figure came from; a public ad library cannot supply it. If the analysis is built on longevity instead, check the actual days-running figure rather than a vague 'it's been running a while.'
| Claim | Value | Source |
|---|---|---|
| What a public ad library can and cannot show | creative, platform, start date, active variant count and destination; no spend, no impressions, no conversions for anyone outside the account | Measured in Real Money, Field Manual |
| Example of longevity as the only real signal available | one $4.99 product in a comparable library ran live for 482 days | Measured in Real Money, Field Manual |