David Baker's White Noise blog has a great post on the lift from multichannel marketing efforts. It describes the "stacking effect", the combined effect of layering multiple channels to influence consumer behavior (either online or offline).
The for-profit education sector enrolls the bulk of its students from calling leads generated via direct marketing - the bulk is the Internet but other channels include TV, outdoor, radio, newspaper, etc. It is tempting to evaluate these channels in silos - for example, a TV ad campaign in a local market might have a ridiculously high cost per lead of $1,000 - but did that particular campaign cause a lift in that market's conversion of Internet leads?
We need to avoid the tempting trap of analyzing advertising sources in a vacuum - it is easier and less time consuming but will likely to poor decisionmaking, as illustrated by an example cited by David - "For a large company, a recent study looked at multiple touchpoints leading to a conversion. When looking at click data alone, we saw that when users clicked on only one channel they had a conversion rate of 4.7%. When they clicked on two channels, their conversion rate almost doubled, to 8.5% -- and when they clicked on three channels, their conversion rate almost tripled, to 11.5%. (Channels included search, display, email, affiliates, portals, and two additional industry-specific channels.)"
