Last Click

Last Click is an attribution model in which one hundred percent of the value of a conversion is attributed to the channel from which the person came the last time before the target action. All previous touches receive a zero.

This is the default model in almost all analytics systems, and this is why it is most often misleading. A person could read you on Telegram for six months, then see an article, then compare it with competitors - and at the last moment enter the brand name into the search. By last click, all the credit will go to branded search, and Telegram and content will turn out to be “useless” channels with zero return.

Hence the typical mistake: they turn off the top part of the funnel because it “doesn’t convert”, and after two months they discover that brand search has also dropped - because the demand that it previously simply collected has stopped appearing.

When last click is appropriate: short transaction cycle, impulse purchases, single-channel acquisition. When it’s harmful: long cycle, multiple touches, large share of content and reach channels. You can check this with a report on associated conversions - if a channel has a lot of them, but there are few last click conversions, the channel is working, just not the last.