Competitor Ad Teardown from Public Ad Libraries

The first thing we look for is duration. An ad still serving after ninety days has survived the only test the public can see, and a teardown that ignores run length is a mood board. A round covers eight to twelve competitors and reports on the arguments they are running, not the assets. Sources are Meta's Ad Library, TikTok Creative Center, Google's Ads Transparency Center, LinkedIn's Ad Library, and the App Store and Play Store creative sets where the product ships an app. What longevity tells you and where it lies: Run length is the only public performance signal that exists, and it misleads in three specific ways. Always-on brand budgets keep weak ads alive indefinitely. Retargeting ads run long on very small spend. And library timestamps record the ad, not the version, so the creative can be swapped underneath a listing that looks continuous. We corroborate rather than trust it. Ads targeted into the EU carry spend ranges and reach in the transparency data. Google's library carries a first-shown date and format. The number of distinct placements and ad sets attached to one creative in a listing says more about conviction than the run length on its own. Where the signals disagree we report the disagreement instead of averaging it away. How the material is organised: By argument, not by format. Forty ads in a category usually resolve into four or five arguments wearing different clothes, and the moment they are grouped that way the picture changes. What comes out is a saturation map: the arguments the category has proven, the ones now crowded enough that your viewer has already heard the sentence this week, and the ones nobody is making. Absence is where the work earns its fee. An argument nobody runs is either an opening or a graveyard, and telling them apart is the judgement you are paying for. Sometimes the argument is absent because it is legally unavailable in the licensed markets. Sometimes it is absent because three operators tried it in 2023 and the library still shows the short run lengths. Sometimes it is genuinely open. We say which, and we say when we are not sure. What you receive: A written teardown per competitor: the arguments they run, the formats carrying each, what their landing page promises against what the ad implied, offer structure and where it sits in the funnel, and hook lines transcribed word for word. A ranked list of arguments with evidence of life attached. A shortlist of gaps, each marked opening or graveyard with the reasoning. Reference stills and clips, every one linked back to its library entry so anyone in your building can check the claim themselves. What we will not do: Build you a swipe file. A folder of competitor ads with no argument attached is the shortest route to making the same ad six weeks later with a worse budget. Recommend cloning a top performer. Recreating a competitor's winner puts you second in a race they started with the audience already warm to their name, and it hands the category their framing. Quote competitor spend as fact. Third-party spend estimates are modelled from panels and impression sampling. We will cite them as a range with the method named, or not at all. Who this is not for: Categories with fewer than about six active advertisers running paid social. There is not enough public material for the reading to be worth what it costs, and we will tell you that on the call rather than after the invoice. Teams who want this monthly as a standing deck. Categories do not turn over fast enough to justify it. Quarterly is the honest cadence for most, and for a live launch window we will watch weekly and say so.