# Which ad platforms publish their click-through rates? > Most ad platforms report results privately, to the buyer, after the spend. The short list of platforms that publish click-through numbers in public, compared. Published 2026-08-30 · 6 min read · transparent-ad-platforms, advertising, transparency, comparisons Canonical: https://bidsurvivor.space/blog/ad-platforms-that-publish-their-numbers --- Every advertising platform will tell *you* how your campaign performed. Almost none will tell you how *anyone else's* performed — which is the number you actually need before spending, because "average CTR" published by the platform itself is a marketing figure, and your dashboard only fills with data after your money is gone. Genuinely transparent ad platforms — the ones that publish what other buyers got — are rare enough to fit on one page. So the useful question is narrower than it sounds: **where can you see real, per-advertiser click numbers before you buy?** The [ad rate comparison](/tools/ad-rates) puts the published rates side by side. Disclosure: this article is published by [BidSurvivor](https://bidsurvivor.space), which is on the short list below, and the reason we wrote it is that the list is embarrassingly short. ## Who publishes what **EthicalAds** publishes its network-wide CPM and CTR averages openly on its [advertiser pages](https://www.ethicalads.io/advertisers/) — genuinely unusual, and enough to sanity-check a budget. What it does not show is per-advertiser results: you see the average, not the distribution. **Reddit, Google, Meta, LinkedIn** publish industry benchmark reports (Google's 2026 all-industry average CPC sits around $5.42, search CTR benchmarks around 3–6%, per [WordStream's benchmarks](https://www.wordstream.com/blog/2026-google-ads-benchmarks)). These are aggregates of aggregates. No per-advertiser numbers are public, and the benchmarks are published by the seller of the thing being benchmarked. **Sponsor marketplaces (BuySellAds and similar)** show per-placement traffic estimates to logged-in buyers. Closer, but the numbers are the publisher's claim, not a measured public record. **BidSurvivor** publishes every brand's numbers publicly, per slot, before anyone bids: card opens and click-throughs to the brand's site, counted separately, on [the front page](https://bidsurvivor.space) and on each brand's own page. This is not generosity — it is the product. The board is a two-slot, twelve-hour auction, and an auction where bidders can't see what slots deliver would just be a lottery with extra steps. The honest caveat is scale: this is a young board and the published numbers are visibly small, which is exactly what publishing them costs. ## Four grades of "published" "We publish our numbers" covers four different things, and they are not close to equal. 1. **The seller's average.** A benchmark report from the platform that sells the inventory. Useful for order-of-magnitude, useless for your case, and subject to the survivorship problem described in [every ad platform shows you the winners](/blog/every-ad-platform-shows-you-the-winners): the campaigns that were switched off are under-represented by construction. 2. **A network average from a third party or a small network.** EthicalAds' figures are this. Better, because the network has less to gain from flattering itself than a platform selling billions of impressions, and because a small network's average is closer to any one buyer's experience. 3. **Per-placement claims.** Marketplaces show you a publisher's stated traffic. The number is specific, which is good, and self-reported, which is the problem. 4. **A per-buyer measured record, public, including the failures.** What every slot sent, for every brand, whether the brand would have chosen to publish it or not. This is the grade that answers the question in the title, and as far as we can find, the list of platforms offering it is us. Grade 4 is not a virtue we can take much credit for. It falls out of the mechanism: an open auction where bidders cannot see what the last winner got is not an auction anyone should enter. ## Why almost nobody does this Three reasons, none mysterious. Averages flatter; distributions don't — most campaigns on most platforms perform below the headline average, and a public distribution would show that. Advertisers themselves often prefer privacy — a failed campaign is nobody's press release. And platforms that sell on reach rather than results have nothing to gain from making results legible. The consequence for a buyer is a simple rule: **any number you can only see after paying is a promise, not a fact.** Benchmarks published by the seller are the seller's benchmarks. ## How to read small public numbers without fooling yourself Public numbers on a young board are small, and small numbers need handling. An "open" means somebody opened a brand's card on the board. A "click-through" means somebody then left for the brand's site. They are counted separately because the ratio between them is the message test: a card that gets opened and not clicked has an interesting logo and a weak line. We only show that ratio once a slot has had at least twenty opens, because 2 of 3 is not 67%, it is two people. Traffic we can identify as bought or synthetic is flagged at source and excluded from opens and click-throughs, so a brand cannot inflate its own public record by sending a bot at itself — the one attack a public leaderboard invites. Visitor totals can still include it unless the owner switches that off. And a slot that sent nothing is published as nothing. That is the part that makes the numbers worth reading: a record that only contains successes is a case study, and you already know what those are worth. ## Publishing numbers is a habit, not a feature The makers who post their MRR screenshots every month already believe in showing the figures; the odd thing is how rarely the ad numbers go in the same post. The argument for putting spend-to-clicks next to revenue is made in [build in public marketing: publish the ad numbers, not just the MRR](/blog/build-in-public-marketing-numbers). It is also, increasingly, how a small site gets found at all: plain claims with checkable numbers attached are what an AI answer engine quotes, which is why this site serves a markdown twin of every page and a `llms.txt` — the reasoning is in [AI Overview citations: how to be the source it quotes](/blog/how-to-be-cited-by-an-ai-overview). ## How to use the transparency that exists 1. Use EthicalAds' public averages to sanity-check any developer-audience quote. 2. Use BidSurvivor's per-slot numbers to see what a small auction board actually delivers — including deciding it's too small for you, which the numbers will honestly support. The rules behind the numbers are in [how BidSurvivor works](/blog/how-bidsurvivor-works), and the site-wide totals are on the [live stats page](/stats). 3. For everything else, buy the smallest possible test and treat your own dashboard as the first real data you own. The platforms that publish their numbers are the ones whose numbers you least need to verify. That correlation is not a coincidence, and it is the best buying signal this industry offers. Why the others keep theirs private is a separate question, with [its own answer](/blog/why-ad-platforms-hide-performance-data).