WebmasterID logoWebmasterID
Attribution models

Cannibalization in measurement

Cannibalization in measurement is the opposite of a halo: a channel captures conversions that another channel — often organic or direct — would have delivered anyway. Branded paid search bidding on terms users would have clicked organically is the canonical case. Attribution credits the paid click, but the incremental value may be small. This page explains cannibalization and how incrementality testing exposes it.

Partially verified

What cannibalization looks like

Cannibalization happens when spend captures demand that already existed and would have converted through a cheaper or free path. The classic example is bidding on your own brand terms: many of those clicks would have arrived via organic results regardless, so the paid conversions overstate incremental value.

It is a measurement convention rather than a single standardized metric; teams define and bound it differently.

Testing for it

Attribution alone cannot reveal cannibalization, because it credits the observed touch regardless of whether that touch was necessary. The fix is an incrementality test: hold out the channel and see whether total conversions actually fall.

If conversions hold steady when the paid channel is paused — with organic or direct rising to compensate — the channel was largely cannibalizing. If they drop, it was incremental. This pairs naturally with the halo effect as the two-sided question of true contribution.

How it appears in analytics and logs

A paid channel with strong attributed conversions but low incremental lift in a holdout is likely cannibalizing organic or direct demand rather than creating new conversions.

Diagnostic use case

Detect when a paid channel is being credited for conversions that organic or direct would have captured for free, overstating the paid channel's incremental value.

What WebmasterID can help detect

WebmasterID's observed organic, direct, and paid session trends let you watch whether a paid push coincides with a drop in free traffic — a cannibalization signal worth testing.

Common mistakes

Privacy and accuracy notes

Cannibalization is assessed with aggregate holdout experiments, not individual tracking. Definitions vary by team; this is educational, not legal advice.

Related pages

Sources and verification notes

Last reviewed 2026-06-24. Facts are checked against primary/official sources where available; uncertain specifics are marked “Data not yet verified” rather than guessed.