How to use the Google Ads Transparency Center for competitor research: what it shows across Search, YouTube and Display, what it hides, and how run length fills the gap.
You can probably recite your main competitor's Meta ads from memory. The UGC angle, the discount hook, the founder testimonial. You check the Meta Ad Library every Monday like half the industry does.
Then one evening you search Google for your own category, and there they are: a Search ad with an offer you have never seen on Meta. Judging by the polish of the copy, it has been running for a while.
Google has had a public ad library since March 2023, announced on Google's own blog as the Ads Transparency Center. Of the three big ad libraries it is the one media buyers use least, and I get why: it hands you the least data per ad. The gaps have a workaround, though, and the parts it does show are the parts competitors guard most carefully on other channels.
The takeaways
Every ad from every verified advertiser, across Google Search, YouTube and the Display network. You search by advertiser name or website at adstransparency.google.com, no login required, then filter by platform, ad format (text, image, video), region and a date range.
Two details earn their keep. First, the verified legal name behind each advertiser: when three lookalike brands crowd your niche, this tells you which company sits behind which storefront. Second, the region filter: pick one country and you see what a competitor serves there and nowhere else. That is how you catch a quiet market entry months before it becomes visible anywhere louder.
One absence to know about: Google's help documentation states that ads only appear once an advertiser completes verification. A brand that never verified is invisible here, so an empty result does not prove an empty Google strategy.
No spend, no impressions, no clicks, no conversions, no keywords, no audience data. The only targeting dimension Google exposes is region. The Transparency Center exists for ad accountability, so it tells you what runs and where, never how well it performs.
That sounds disqualifying, and for some jobs it is. You cannot reconstruct a competitor's budget split from it, and you cannot see which query triggered a Search ad. What you can read is the message itself: the exact headlines, the offers, the video creatives a competitor is paying to run today. For creative and angle research, the message is most of the value. Performance you can infer from the outside, which brings me to the workaround.
By run length. Set the date filter to a window two or three months back, note the creatives, then compare against what serves today. Anything alive in both windows has survived 60+ days of someone else's optimization, and direct-response advertisers do not carry losers that long.
I treat 30 days as the threshold. The competitor-intelligence side of Adscalr, the tool I built, applies that same rule across all three ad libraries it pulls daily: an ad still running after 30+ days gets flagged as a durable winner, because longevity is the one signal a library cannot strip out. It is also the same evidence standard I hold my own tests to, where a hot first week proves very little.
Search ads carry the highest-stakes copy a competitor writes. Google caps a Search headline at 30 characters, and those characters face people who are about to buy. Whatever made it in there, the guarantee, the price anchor, the objection answered first, is the claim that competitor trusts with its priciest clicks.
YouTube is the hook library. Filter to video, find what has run longest, then watch the first five seconds of each. Every one of those openings survived a skip button, which is a harsher test than any feed scroll.
Display is the retreat. Banners do the retargeting work, so they expose the fallback offer: the discount a competitor is willing to hand a visitor who already walked away once.
The Meta Ad Library shows you a competitor's demand creation: how they interrupt a feed. The Transparency Center shows their demand capture: what they say when a buyer is already searching, and what they run on YouTube while everyone watches Meta. A competitor who looks quiet in the Meta Ad Library can be loud on Google, and from Meta alone you would never know.
I checked all of this by hand for years, three tabs every Monday morning. Eventually that became one of the reasons I automated it: Adscalr pulls Google Ads Transparency daily alongside the Meta and TikTok ad libraries, normalizes everything into one dataset, and a vision AI decodes each static into around 20 structured fields. Google contributes the thinnest metadata of the three, which is exactly where the durability lens earns its keep. If reading three libraries against each other sounds like work you would rather not repeat weekly, that is what the competitor intelligence side of the product exists for.
This is the thinking behind Adscalr.
See the product →