Industry analysis and the case for advertiser-side attribution.
DAP computes aggregate statistics privately. Croupier produces per-conversion proofs. Different questions, different architectures, different tradeoffs.
Every cookie replacement recreates the same problems. Bearer attribution never depended on cookies in the first place.
DoubleVerify missed 21% of bots. IAS labeled bot traffic as human 77% of the time. Verification asks the wrong question.
Google lost its second antitrust case. Divestiture of AdX and DFP may follow. What this means for measurement independence.
Creator ad spend hit $37B but 57% of marketers can’t track ROI. SAG-AFTRA is organizing creators. The real issue is provability.
From Google’s monopoly ruling to phantom conversions, ten evidence-backed reasons platform self-reporting cannot be trusted.
Internal docs show 19% of Meta’s China ad revenue came from scams. They had an anti-fraud team. They disbanded it.
A comprehensive overview of attribution models, what broke in 2026, and where the industry is heading.
A niche podcast at 12% conversion beats a content farm at 0.01%. Cryptographic coupons give publishers the receipts to prove it.
Triple Whale relies on forgeable UTM parameters and modeled attribution; Croupier proves what actually happened.
Forbes ran a secret subdomain as a content farm. $770M per quarter goes to MFA sites. Coupons that need real customers expose the fakes.
Northbeam uses black-box ML to assign fractional credit; Croupier provides cryptographic proof of each touchpoint.
Honey wasn’t alone. Capital One Shopping and Rakuten did the same thing. The cookie is the vulnerability, not the extension.
Privacy Sandbox gave attribution control to the browser vendor that is also the largest ad platform.
Platform-reported conversions routinely exceed actual sales because every platform claims credit for the same event.
DoubleVerify checks viewability and bot traffic but cannot tell you which impression caused a conversion.
Signed tokens instead of cookies. The advertiser issues, the customer carries, the advertiser verifies. No platform dashboard needed.
AppsFlyer depends on self-reported data from the ad networks it is supposed to measure.
UTM parameters are unsigned plain-text strings anyone can forge; Croupier replaces them with cryptographically signed tokens.
Platform-reported metrics inflate conversions by 30–50%. Here’s why the numbers don’t add up and what independent measurement looks like.
Browser extensions and affiliates silently overwrite tracking cookies, misattributing conversions at scale.
MegaLag exposed PayPal’s Honey overwriting affiliate cookies at checkout. 25+ lawsuits followed. Here’s what it means for attribution.