Glossary
Marketing attribution is the practice of assigning credit for a conversion to the marketing interactions that preceded it, so spending decisions can be based on contribution rather than on whichever channel happened to be last.
Someone reads an article, searches the brand a week later, clicks an ad, then arrives directly and enquires. Crediting the enquiry entirely to the final click describes the last step, not the reason.
Last-click over-credits capture channels such as brand search. First-click over-credits discovery. Linear treats a passing visit as equal to a decisive one. Data-driven models are better but need volume, and remain estimates rather than measurements.
Privacy controls, cross-device journeys and searches that end without a click all remove signal. Attribution now describes a decreasing share of what actually happened, and treating it as complete is a mistake.
Whichever you understand the bias of, applied consistently. Comparing the same decision under two models is more informative than trusting one. Switching models to make a channel look better is how attribution stops being useful.
Because attribution only sees tracked interactions. Privacy controls, cross-device journeys and zero-click searches all remove signal, so the attributed figure is a subset of reality, not a reconciliation of it.
Marketing attribution is the practice of assigning credit for a conversion to the marketing interactions that preceded it, so spending decisions can be based on contribution rather than on whichever channel happened to be last.
Connected
