whatismmm

MMM vs attribution

Attribution follows individual users across touchpoints and divides credit among the ads they saw. MMM ignores individuals entirely and estimates channel effects from aggregate spend and revenue history. Attribution needs user-level tracking, which privacy changes gutted; MMM needs history and variation, which nobody can revoke.

By Oliver Wakefield-Smith · Updated 2 August 2026 · Every number on this page resolves to a primary source in the ledger below.

Different units of analysis, not different opinions

Multi-touch attribution's unit is the user journey: this person saw a video ad, clicked a search ad, bought; divide the credit. Its promise is granularity down to the campaign and creative. Its dependency is observing the journey, which requires identifiers that persist across apps and sites.

MMM's unit is the week (or the week-by-region cell). It never asks who saw what; it asks whether revenue moved when spend moved, controlling for everything else it can observe. The cost of that abstraction is granularity: an MMM tells you what search did, not what the Tuesday exact-match campaign did.

What iOS actually broke

Apple's App Tracking Transparency framework, introduced with iOS 14.5, requires explicit opt-in before an app may track users across other companies' apps and websites.1 Cross-app identifiers thinned, journeys fragmented, and click-based attribution lost coverage precisely where mobile-first brands spent. The platforms responded with modeled conversions, which is attribution filling its gaps with statistics, at which point the honest question became: if the dashboard is partly a model anyway, why not use the model class built for the job? That is the short version of why MMM returned to fashion, and why Google and Meta both released open-source MMM frameworks.2

When each is the right tool

Attribution, or what remains of it, is still the right instrument for tactical optimization inside a platform: creative tests, audience comparisons, bid changes, where relative signal matters and biases roughly cancel. MMM is the right instrument for cross-channel allocation and for any number that leaves the marketing department, because its estimates are incremental by construction.

Can they disagree without one being wrong? Routinely. Platform attribution reports touched revenue; MMM reports caused revenue. A retargeting ROAS of 10 and an MMM estimate of 1.5 for the same channel can both be accurate answers to their respective questions; only one of those questions is worth budget. The reconciliation discipline is triangulation.

Source ledger

Every numeric claim on this page resolves to one of these primary sources.

  1. [1]Apple: App Store user privacy and data use (App Tracking Transparency)https://developer.apple.com/app-store/user-privacy-and-data-use/retrieved 2026-08-02
  2. [2]Google Meridian documentationhttps://developers.google.com/meridianretrieved 2026-08-02