Cost to serve modelling

Many of our company's clients (plus anyone who has cornered me on the subject) knows that I can talk for ages on the subject of Cost-to-serve models. But for those who haven't (and apparently you should count yourselves lucky), here's my main theory:

The average cost to serve a customer (e.g. to make a sale or to answer a complaint) depends upon the channel they use to contact you and then how you deal with that contact. Basically, some channels (e.g. face-to-face) cost lots and others cost less. Companies wishing to minimise their costs, should therefore try to 'manage' customers to the lowest relevant cost channel required to provide the level of interaction needed to service the cuctomer.

For example, in the diagram below 'B' shows that a high level of customer interaction (e.g. face-to-face contact to sell a new car) has a higher cost to serve than the lower 'A' which needs lesser levels of interaction (e.g. a web page). Basically, the more complex or 'human' you need the service to be, the more it costs to achieve.



The difficulty comes when a company badly manages their customers down to lower cost to serve channels, especially when the customer will always try to use the one with the highest possible level of Interaction.