The Purchase Funnel is a variant of communication theories like AIDA (‘attention’, ‘interest’, ‘desire’, ‘action’) or DAGMAR* which became popular in the 1950′s and ’60′s. These models were derived from classical learning theory and have in common the assumption that, for marketing communications to be effective, the consumer must go through a linear sequence of mental events, with each stage dependent on the success of the previous stage, producing a ladder-like progression towards purchase.

Although developed half a century ago when the world was a very different place, linear sequential models such as the Purchase Funnel still provide the basic metaphor and language of marketers today because the logic of the sequence seems irrefutable and the supposed stages lend themselves readily to objective measurement.

It all seems reasonable enough. So what’s the problem?

The problem is that it doesn’t even come close to mapping the consumer’s approach to a considered purchase decision in the Internet age. In the automotive category, for example, the 2004 New Autoshopper.com Study** found that among all new-vehicle buyers, approximately 50% say their make/model decision and the price paid/offered were impacted by automotive information found on the Internet—up from about 40% in 2002. The Purchase Funnel takes no account of the way consumers use online research to 1) expand their consideration set and 2) to take advantage of the experience of existing owners and users to help guide their choice. Both points have important implications for the purchase funnel.
________________________________________
* Defining Advertising Goals for Measured Results, 1961 (Russell Colley)
** Based on responses from 26,838 consumers who leased or purchased a new vehicle registered in January or February 2004

How so? Let’s take both points separately.

Expansion of the Consideration Set

A recent study of search behavior online*** revealed that generic terms account for the majority of online search activity associated with a subsequent purchase. The study of search behavior in the consumer electronics category showed that generic product search terms (e.g. “camera,” “plasma television” or “PDA phone”) accounted for more than 70% of total search volume, while trademarked retailer terms (e.g. “Best Buy,” “Gateway.com”) accounted for 20% and specific product terms (e.g. “Canon digital camcorder,” “HP notebook nx9010″) accounted for 10%.

It also found that while 85% of searchers do indeed conduct additional CE/C searches later in the shopping process, the majority of consumers continue to use the same search term type (either generic or branded) with which they began the search process. As importantly, comScore also found that generic search terms are likely to have influenced even those consumers who converted to purchase after conducting a retailer trademark search (e.g. “Best Buy” or “Gateway”). Fully 84% of these buyers searched using a generic term earlier in the buying cycle, reinforcing the importance of reaching consumers early in the search process when they are defining their consideration set.

The implications of this study challenge a widely held assumption encoded in the purchase funnel, that most consumers begin the product search process by using a generic search term (e.g. “plasma TV”) and then later refine their search activity to product-specific terms (e.g. “Sony Plasma KE-42M1″). It also suggests that the use of generic search terms will invariably open consumers up to the possibility of being tempted by new brands and products that were not in their original consideration set.

And, in fact, we already see this happening in the automotive category. CNW Research is considered the industry standard measurement of consumers’ automotive purchase behavior. Historically, it has reported that consumers start out with six or seven vehicle models in their consideration set and then, over the following six months or so, gradually refine the list down to the one or two vehicles that they will test drive. The latest annual tracking studies from CNW reveal that around half way through this process, the number of models investigated jumps back up from around four, to five or six. We believe this spike is reflecting new make and model considerations generated by online searches for information about their original considered set.

We also believe that what influences consumers most in this stage is not what the manufacturer says about the ownership experience but what current owners say, which brings us to point number two.
_________________________________
*** This ComScore study is based on a massive, global cross-section of more than 2 million consumers who have given ComScore explicit permission to confidentially capture their browsing and transaction behavior, including online and offline purchasing.