So SEW (and several other folks) pointed to a study today from Comscore & Google that shows a very high percentage of Internet surfing results in a latent, offline purchase.  From Chris Sherman:

...25% of product-related searches resulted in purchases, with 63% occurring in brick and mortar retailers, compared with 37% from online merchants. The study has similar findings to a Yahoo-sponsored ComScore study in December 2004, which found that 92 percent of purchases related to consumer electronics occurred offline...

For those marketers who measure conversion rates only on the web, a vast chasm exists between tracking and actual results. So how do you measure offline conversions? There's no cookie you can attach to a human being or their credit card and asking salespeople (whether it's B2B transactions or consumer activity) to collect the information provides lackluster results.

The methodology has to lie in comparitive analysis - checking sales figures for Internet promotions and visits against increases in offline sales. With the proper tracking in place, you can see side-by-side comparisons of visitor analytics showing traffic to a particular page (or section) and analyze sales figures over time to determine where the increases are coming from.

Offline tracking, however, provides intangible data and if done consistently, can be especially time consuming and expensive. Hence, it becomes wise to eventually cheat :)

By cheating, I'm referring to using previous data to predict and measure new trends. For example, if you already know that putting a Seagate hard drive on your homepage results in a 5% click-through to that product and boosts in-store sales by 5%, you can go ahead and put a new product there, measure click-throughs and predict what in-store sales boost you'll get. Don't forget to refine and test your analytics system, but this "poor man's" method can be handy in a pinch.

Anyone out there run online campaigns and track offline conversions?