Okay, I started this out with a lie, this isn't a new way to sell at all. The sense and respond approach to selling is as old as business itself. Think about an old fashion "mom and pop" business, it operated by knowing its customers intimately. This intimacy came from both the type of interaction and the frequency. In this older model of community it was not difficult to profile each customer, you knew them, their family, friends and much of their personal activities. In our "modern" world, we've lost that level of intimacy with shopkeepers and suppliers. This unique sense of community still exists, but its very rare today...or is it?
Online communities mimic offline counterparts in many ways. What they remove is the need for members of a community to be geographically close. Proximity is replaced with the Internet, which fundamentally substitutes one type of relationship building with another. More and more people are living their lives out online and providing a great deal of information about themselves in the process. This information goes way beyond transaction data, and could provide a very complete social profile if assembled and analyzed. I should take a minute to remind everyone that this is not creepy in any way, the information I'm talking about is freely provided and publicly available.
Now it is possible to go back to a sense and respond sales model. As a business you track transaction data, which provides one view of individual behavior. If you then collect, correctly identify, and map social online data to this transaction data, you have the foundation of a very rich understanding of your customer. Sensing (or listening, collecting, etc.) is not technically difficult, there are many listening tools available. The part that is somewhat more difficult is correctly identifying the individual to whom the data belongs and matching that to the customer data you already collect. This issue of knowing your customer wherever you interact with him/her is a critical problem for most businesses today. Businesses have taken a "channel" approach to sales, that is, tracking customers effectively but only through a single channel or at best through 2 channels. The last part of the process, mapping the social data to transaction data is also not difficult, assuming you have identified the customer correctly.
Companies are moving beyond channels, or even cross-channels and into a model that must be flexible across any touch point. There is a need then for a sort of omni channel identity, or the ability to correctly identify customers whenever and where ever you interact with them. This one "simple" requirement is a serious challenge for CRM systems and companies and until it's solved, will hamper the overall customer experience. Loyalty programs can help get customers to self identify, but are not yet at least, enough.
Needs based selling, or sensing what the customer needs and providing it is a very powerful sales tool. To take that one step more though, could you build a behavioral model that provides the clues to actually predict what a customer needs, maybe even before the customer clearly understands that need themselves? The next generation of sales is predictive and the data that's becoming available could make that a reality.