Thoughtful MCM article by Cam Balzer on quantifying the value of off-web conversions resulting from paid search:
In one survey, we polled the client's online customers as well as an anonymous panel regarding their shopping and purchase behavior online and offline. From that information and purchase data we were able to determine that the client's customer base exhibited an offline multiplier of $2.50: For every dollar in online sales generated by search, an additional $2.50 is generated offline.
This multiplier will vary by merchant, and can be considerably higher or lower. For example, the multiplier for storeless pure-plays like Amazon is, by definition, zero.
My instinct is that for most mechants, $2.50 is too high -- I suspect $0.50 is more typical. Regardless, if you have stores, your multipler isn't zero.
Surveys are one way to get at this number for your business, though there's often material differences between what people report and what they actually do. Online coupons are another way, if the coupon can be scanned at the store register and later tracked without hassle. Cam didn't discuss call centers, which also pull orders out of the offline channel. One powerful way to track call-center spillover is to have your call agents ask for a visible tracking code on the web page.
Estimating this multiplier correctly is important. Set it too low, and you leave opportunity on the table, ceding share to your competition. Set it too high, and you advertise recklessly, burning ad budget needlessly. However, when you see smart multi-channel competitors consistently outbidding you in search, it is time to wonder if they place more value on channel spillover than you do, and it may be time to put some serious effort into an accurate estimate of your off-web multiplier.
After writing this, my colleague George Michie pointed out the importance of distinguishing research from shopping. A consumer moving towards a big-ticket considered purchase -- say, an expensive piece of consumer electronics -- will use the search engines heavily during the research phase. If later polled at point of purchase, many will report having used search to reach the retailer's site prior to buying. A careful survey or well-designed coupon approach should distinguish between search visitors who come to your site just for research, versus those who come to buy and then opt to do so off-line.