Where does the credit for a TV campaign actually land in your analytics?

Where does the credit for a TV campaign actually land in your analytics?

A brand ran national TV, then watched branded search and direct traffic spike. Last-click handed the credit to search and "direct." The question: did search earn those conversions, or did TV manufacture the demand search merely caught?

The study type. Several media mix modeling and matched-market analyses (including widely cited work from Analytic Partners and platform incrementality teams) have isolated the "TV halo" — the lift TV drives in lower-funnel channels — by comparing markets that received heavy TV against matched markets that didn't, then attributing the search and direct delta to TV.

What they found. In documented matched-market cases, 20-40% of the search and direct conversions occurring during flights were incrementally caused by TV, not by the search channel that recorded them. One frequently cited result: pausing TV in test markets dropped branded search volume measurably within days, with no other change — direct causal evidence of the halo.

The nuance. This is the most common attribution error in disguise. The downstream channel correlates perfectly with conversion because it sits closest to it. The upstream channel that created the intent gets nothing in click-based models. Correlation lands the credit; causation lives upstream.

Bottom line for practitioners: when an upper-funnel channel runs, your last-click "direct" and "branded search" lines will rise and steal its credit. Use matched-market holdouts to measure the halo — pause the upstream channel and watch the downstream channel deflate. The 20-40% that vanishes was never search's to claim.