Content is systematically undercredited by last-click attribution, because it does its work early and gets no credit at the close. With consent-gated analytics and dark social on top, the standard dashboard now actively misleads.
Why the default number is wrong
A person reads three articles over two months, later searches your brand name, and converts. Last-click credits branded search. The content that created the demand shows nothing. Defund it on that basis and pipeline dries up a quarter later with no visible cause.
Three measurements that hold up
Self-reported attribution. A required how did you hear about us field with a free-text option on your demo form. Imprecise and recency-biased, and still better than a model that cannot see a large share of your traffic.
Assisted view over a long window. Set the lookback to ninety days or more and look at assisted conversions rather than last click. It does not fix consent gaps but stops crediting the final touch for everything.
Cohort comparison. Compare conversion rate and deal size between people who read three or more articles before first contact and those who read none. The gap is the honest content effect, and it is usually large enough to end the argument.
The cost side
Include salaries. Content teams routinely compare a paid CAC that includes media spend against a content CAC that excludes the people producing it. Use the same cost basis on both sides or the comparison is theatre.
What to report upward
Blended CAC alongside a written log of what changed in the mix each month. You lose per-channel precision and gain a number nobody can game by adjusting attribution settings.
General information, not legal advice. Rules current as of July 2026.