What Changed in Programmatic Attribution Since 2023

This page used to carry a long-form version of an argument that now lives, in full, at a single canonical address. If you came here for why affiliate tracking and click attribution fall short in programmatic, read that instead — it is the complete and maintained version.

Read this instead: Why affiliate tracking and click attribution fall short in programmatic advertising

Why this page is short now

Two pages on this site covered the same question under near-identical titles. That splits the evidence, splits any authority the argument earns, and leaves a reader guessing which one is current. So one of them is canonical and this one points at it.

The canonical version was chosen on the cleaner address, the longer history and the fact that every figure in it is ThumbAd's own published measurement rather than a third-party statistic. That matters for a page whose whole subject is whether numbers can be trusted.

What changed since 2023, in brief

The original article, published in April 2023, argued that affiliate tracking and click attribution misread programmatic and that Google Analytics could not close the gap. Three years on, that has held up better than most predictions of the period.

• The third-party cookie did not disappear on the timetable the industry planned around. Building measurement that depends on it is still a bet on somebody else's roadmap.

• GA4's event model improved cross-channel reporting without moving the boundary that matters: an impression that never touched your site cannot be observed by a tool that only watches your site.

• The centre of gravity moved toward first-party data and toward reconciliation against systems the operator owns — which is where it should have been all along.

None of that changes the conclusion. It sharpens it.

The evidence, in one line

Across six independent measurements — five operators, two continents, six years — the post-view share of conversions landed between 89% and 96.2%. Every one was customer-ID-level or platform-verified; none was modelled. An operator reading clicks alone is seeing roughly one conversion in ten. That is a measurement-completeness finding, not an incrementality claim.

The method sits in click and view attribution in programmatic advertising, and what it costs when nobody reconciles is documented in The 2% Problem.

To test any desk against ThumbAd's published due-diligence standard — including ThumbAd — the 20-question partner scorecard scores it against a bar of 24 out of 30, in your browser, with no email required.

ThumbAd is an independent programmatic trading desk in Oslo, trading on The Trade Desk across Europe, Africa, North America and Latin America since 2012, with betting and gaming at the core.