With customers increasingly using social tools (Facebook tops 500 million users with 41% of Americans having a Facebook page, 7% of Americans use Twitter, and YouTube gets 2 billion views per day), how do you interact with and service your customers where they are spending their time?

Let’s focus on Facebook, as it has the potential to evolve into a shopping and service destination for retailers. Two support models are evolving. One model is to engage in support activities via a separate tab on a Facebook company page. Once a user clicks on this tab, they can engage with a community of peers or a customer service agent without leaving the site. There are vendors, like RightNow, Parature, Genesys, and Get Satisfaction, that offer apps that do just this.

The other Facebook support model that is emerging is one that is based on listening to all conversations happening on a wall, separating the noise from actionable posts, then routing actionable inquiries to a customer service department so that an agent can respond directly to the post — either on Facebook or by engaging the customer over a more appropriate channel — for example, like pushing a chat link to them on Facebook. This is the model that salesforce.com  and LivePerson are advocating.

 These customer service models are fundamentally different, even though they accomplish the same goal of servicing a customer. Both models offer transparency and visibility into customer service issues being reported by customers. However, one model gives users a clearly defined playground to engage in — a support tab where you can access a peer-to-peer forum, or a corporate knowledge base, or a customer service representative who is separate from the sales or marketing content. The other model is more collaborative and fluid, where the walls between information pushed from the company and questions or comments that users may have about products and services are intermingled. There is no artificial delineation among sales, marketing, and customer service activities. It can also be used to deliver more contextual service based on usage patterns.

I’m not sure which model will win in the eyes of the customer, but I would be very interested in hearing your opinions.