<- all tokdocs
2026.10.03@reincurtoyfule575

Meta's Ad Library Never Shows Commercial Ad Spend, In The UK Or Anywhere Else

Marketing
Watch on TikTok

View on TikTok ->

The artifact is a 96.41 second clip (1:36), 1080x1920 portrait at 30 fps, HEVC Main video at roughly 153 kbps alongside HE-AACv2 audio at 48 kbps and 44.1 kHz stereo, 2,501,383 bytes on disk for an overall bitrate near 208 kbps. It was posted to TikTok on 2026-10-03 at 20:22:35 UTC by handle @reincurtoyfule575 under numeric uploader id 7298143886525301802, with channel nickname reincurtoyfule575. At capture on 2026-10-06 it carried 12,500 views, 775 likes, 5 comments and 394 reposts. The audio track is credited as "original sound" by reincurtoyfule575, which is to say a direct voiceover with no licensed music bed. I read 18 of the 48 extracted frames plus eight higher-resolution crops pulled from the source mp4 at 52, 56, 60, 66, 70, 76, 80, 84, 88, 90, 94 and 95 seconds, against a 379 word transcript. A red rounded caption plate sits at the top of every frame in white bold type: "The 100 Percent Free Reverse Engineering Toolkit For Founders" followed by a hammer-and-wrench and a lightbulb glyph. The first six frames are a talking head, a man in a black jacket over a white t-shirt with AirPods in and a lavalier mic clipped at his chest, on a rust-orange couch against a white wall. From roughly the twelfth second it cuts to a screen share with his webcam reduced to a panel at lower left, over a dark chalkboard backdrop bearing faint words including "SUCCESS" and "IDEAS". The browser shows the Meta Ad Library: a Meta wordmark top left, nav links reading "Ad Library", "Ad Library Report" and "Ad Library API" top right, a "Search ads" box, and panels titled "Explore other tools", "Download information and track spending" with a "Go to Report" button, "Customize your own search" with a "Go to API" button, and "Search for branded content". He sets the country dropdown to "United States", the category dropdown to "All ads", and types "marketing" into the search field, triggering an autocomplete panel headed "Related searches" and "Advertisers". The results screen reads ">50,000 results" above the line "These results include ads that match your keyword search.", with a grouping header "Launched June 2025". Cards carry a green "Active" pill, "Library ID: 1024581229760708", a start line reading "Started running on Jun 23, 2025" or "Started running on Jun 24, 2025 · Total active time 18 hrs", a "Platforms" icon row, and a "See ad details" button. Visible advertisers include Cabrillo College, Benjamin Golden across five adjacent cards, Stanly Community College SBC and Gunly. He then opens a "Filters" modal with rows for "Advertiser", "Platform", "Media type", "Active status" set to "Active ads", and an "Impressions by date" block captioned "Choose an exact date or select a date range" with "From" prefilled to "Jun 25, 2018" and "To" prefilled to "6/25/2025", a date picker open on "June 2025" which he pages back to "March 2025". After applying, the page reloads to "~9,000 results" under the header "Launched March 2025", with filter chips reading "Active status: Active ads" and "Impressions by date: On or before Ma...", and every visible card reading "Started running on Mar 1, 2025". The final frames show him hovering a "Paving Marketers" card, which pops a Page panel reading "Marketing Agency / THE #1 PAVING MARKETING AGENCY We know how to get leads and estimates for paving companies - with a proven track record of doing it. Web, SEO, PPC, Social Media", with "318 likes" and "View ads" and "Go to Page" buttons.

The clip shows one tool, and the screen recording predates the TikTok post by fifteen months

The hook promises a "toolkit" and the caption says "100 Percent Free". In 96 seconds the video names one thing: the Meta Ad Library. He says "the first tool I'm going to show you is called the meta ads library, as in the Facebook ads library" and never reaches a second. The clip ends mid-demonstration on the line "All of these marketing ads have been running for months." It is a fragment cut from a longer piece, which matters for anyone trying to evaluate a "100% free" claim across tools that are not in the file.

The dating is also off. He says "Right now it's late July." The screen disagrees. The "Impressions by date" field prefills "To" with the current date, and on screen it reads "6/25/2025". The picker opens on June 2025, the group header reads "Launched June 2025", and the newest cards read "Started running on Jun 23, 2025" through "Jun 25, 2025". The capture was made on or about 25 June 2025. Whatever the narration was recorded against, the screen recording is a different session, and both predate the 2026-10-03 TikTok post by more than a year. The posting account carries an auto-generated handle and no original audio beyond this voiceover.

The UK spend claim is wrong in three separate ways

This is the load-bearing factual claim in the clip and it does not survive contact with the sources. He says: "if you look at the Facebook ads library in the UK, you can even see how much money is being spent on those ads because of disclosure laws that exist in Europe that don't exist over here in the US."

First, spend is not published for commercial ads in any country. Meta's own Ad Library tools page states that ads about social issues, elections or politics carry "additional information, including spend, reach and funding entities." That additional information is scoped to that category. The marketing ads he is pointing at on screen, the Shopify and med-spa and paving-company ads, are commercial. Pulling up the same search with the country set to United Kingdom returns the same fields it returns for the United States: library ID, start date, platforms, active status and the creative. No budget figure appears.

Second, the United States discloses political ad spend. Meta's documentation says the Ad Library Report "is available in all countries that require authorization to run ads about social issues, elections or politics", and the United States is one of them. US political ad spend ranges have been in the Ad Library since 2018. The asymmetry he describes, where Europe shows spend and the US does not, is backwards about the US half as well as the UK half.

Third, the United Kingdom is not in the European Union and the EU regime he is gesturing at does not apply there. The relevant EU rule is the Digital Services Act. Article 39(2) lists exactly what a very large online platform's ad repository must contain: the content of the advertisement, the person on whose behalf it is presented, the person who paid for it if different, the period during which it was presented, the targeting parameters used, and "the total number of recipients of the service reached and, where applicable, aggregate numbers broken down by Member State." Advertising spend is absent from that list. Meta's August 2023 DSA announcement describes what it added to comply: "expanding our Ad Library to display and archive all ads that target people in the EU, along with dates the ad ran, the parameters used for targeting", stored for a year. Targeting parameters and reach, not money.

There is a real EU advantage under the mistake, worth stating correctly because it is useful. For any ad delivered in the EU, commercial ads included, the Ad Library exposes the targeting parameters the advertiser chose and the aggregate reach. That tells a founder who the advertiser thinks their buyer is, which is more actionable than a spend bracket. DSA retention runs one year from last impression, against seven years for political ads.

One more thing has changed since the recording. Following the EU's Transparency and Targeting of Political Advertising regulation, Meta stopped running political, electoral and social issue ads in the EU as of 6 October 2025. The EU political spend archive still exists for historical ads, and nothing new is being added to it.

The filter he demonstrates does work, and his own numbers quantify it

The technique is sound and the clip documents it well enough to reproduce. He opens Filters, leaves "Active status" on "Active ads", and sets the "Impressions by date" To field backwards, producing the chip "Impressions by date: On or before Mar 1, 2025". Those two conditions together return ads that had delivered at least one impression by 1 March 2025 and are still in delivery, which is a usable proxy for longevity. He says "let's say four months ago. We'll say March 1st", which from 25 June 2025 is just under four months.

His own on-screen counts give the magnitude. The unfiltered "marketing" search in the United States returns ">50,000 results". After the filter it returns "~9,000 results". That is roughly an 82 percent reduction, and it is the single most useful number in the video. The default "From" date of "Jun 25, 2018" is the other quiet detail worth noticing: it sits exactly seven years before the capture date, which lines up with Meta's seven-year retention window for ads about social issues, elections or politics.

"Copy the ad that has run for months" breaks on the evidence in his own screenshot

He is unambiguous about the heuristic: "if you see an ad that has been up for months, then it is a banger ad and you should copy everything about that ad and change as little as possible."

Two fields in his own capture undercut the inference. The June cards display "Started running on Jun 24, 2025 · Total active time 18 hrs" as two separate values, so Meta tracks elapsed delivery time independently of the start date. A library entry with an old start date and a small active time has been paused and restarted. I could not find Meta documentation defining how "Total active time" is computed, so treat the exact semantics as unverified. The existence of two distinct fields is enough to show that "started running on" alone does not establish continuous delivery.

The second field is plainer. Several cards in the March 2025 result set carry the line "This ad has multiple versions" with an info icon, including the Stanly Community College SBC card and one of the Wilsons Auto Sales LLC cards. A single library entry can hold a changing set of creatives. An entry that has "been up for months" may be a slot the advertiser keeps refreshing, which is the exact behaviour he describes two sentences earlier when he says "the name of the game is creatives, refreshing creatives." The library entry persisting is not the same evidence as one creative winning.

The long-running ad he is implicitly endorsing contains a false statistic

The five adjacent Benjamin Golden cards dominate his first result page, and their copy is legible in the high-resolution crops. One card reads: "Adding this ChatGPT code to our Shopify store added $22,120 in additional revenue, all with free AI traffic. Let's face it, SEO is changing. April 2025 is the first time Google's traffic share has dropped under 90%, since 2015."

That last sentence is false. Statcounter data reported by Search Engine Land in January 2025 put Google's global search market share below 90 percent for the final three months of 2024, at 89.34 percent in October, 89.99 percent in November and 89.73 percent in December. 9to5Google's 15 January 2025 write-up corroborates the same Statcounter-sourced finding, that Google "has dropped below 90% share in the worldwide search engine market over the past three months." The threshold was crossed in October 2024, roughly six months before the date the ad claims. Search Engine Land returned HTTP 403 to a direct fetch, so the monthly figures come from its indexed summary rather than a page I rendered myself; the October 2024 timing is confirmed independently by 9to5Google.

This is the practical problem with "copy everything about that ad and change as little as possible." The method ranks ads by survival in the library, and survival measures the advertiser's willingness to keep paying. It does not measure whether the claims in the creative are true. A founder who copies the Benjamin Golden structure wholesale inherits a verifiably wrong statistic about Google's market share, and attaches it to their own brand.

Key Takeaways

  • Correction: The video says UK ad spend is visible "because of disclosure laws that exist in Europe that don't exist over here in the US." Meta publishes spend only for ads about social issues, elections or politics, in every country where those ads require authorization, including the United States. Commercial ad spend is published nowhere.
  • Correction: The UK is outside the EU, so the Digital Services Act does not apply there. DSA Article 39(2) requires content, advertiser, payer, run period, targeting parameters and total recipients reached. It does not require spend.
  • Correction: The on-screen ad the video holds up as a long-running winner states "April 2025 is the first time Google's traffic share has dropped under 90%, since 2015." Statcounter data put Google below 90 percent in October, November and December 2024, as reported in January 2025.
  • Correction: He says "Right now it's late July." The Ad Library prefills "To" with 6/25/2025, the picker opens on June 2025, and cards read Jun 23 through Jun 25, 2025. The capture is from 25 June 2025, and the clip was posted 2026-10-03.
  • Partial correction: The caption promises a "toolkit" that is "100% free". Only the Meta Ad Library appears in the 96 seconds. It has no paid tier.
  • Verified: The filter he demonstrates works as shown. "Active status: Active ads" combined with "Impressions by date: On or before Mar 1, 2025" cut his "marketing" search from ">50,000 results" to "~9,000 results", roughly an 82 percent reduction, per his own on-screen counts.
  • Context: The real EU-only advantage is targeting parameters and aggregate reach for all EU-delivered ads, commercial ones included, added for DSA compliance in August 2023 and retained for one year.
  • Context: Meta stopped running political, electoral and social issue ads in the EU on 6 October 2025 under the TTPA regulation, so the EU political spend archive is now historical.
  • Unverified: Meta's "Total active time" field appears on screen as "18 hrs" against a Jun 24, 2025 start date, but I found no Meta documentation defining how it is computed. I checked Meta's transparency research tools page and searched for help-centre coverage without finding a definition.
  • Unverified: Whether the Ad Library requires a login is contested. Meta's transparency research tools page reads as though sign-in is required; multiple third-party guides state commercial ads load without an account and only political ads are gated. Direct fetches of facebook.com/ads/library were refused, so I could not confirm either way.

Resources

Published October 3, 2026. Writeup generated from a favorited TikTok.