While organizations start with some basic understanding about their target customers, the target segment gets refined with time and may undergo significant changes once there is visibility of ground realities. In organizations that have implemented CRM solutions, the CRM application tracks information about customers across different channels. Overlaying analytics on top of CRM data, organizations can identify different segments, outlining buying habits like preferred products/services, channel, location, willingness to pay for premium services etc. A successful segmentation strategy will result in segments that are easily definable, sizable, reachable and actionable. Segments are defined in such a way that characteristics within a segment are homogenous while between segments are heterogeneous, so that different treatments can be designed for different segments. Characteristics, based on which customers can be segmented can belong to any of the four major types:
- Geographic (e.g., country, region, climate)
- Demographic (e.g., age, sex, income, education, # household members)
- Psychographic (e.g., lifestyle, personality, values)
- Behavioral (e.g., usage rate, loyalty status, usage occasion)
Geographic variables provide high level indicators. Mobile telephony customers based out of coastal regions, where one of the primary occupations is fishing, have a higher willingness to pay for weather related services and forecasts on their phones.
Demographic variables can indicate if a particular age group uses specific types of services more often than others. SMS is used more often in the age group 15-25 than in any other age group.
Psychographic variables are based on human psychology. Customers can be identified as Innovators, Adopters or Followers. High end new products or services can be targeted specifically at Innovators thereby increasing acceptance rates.
Behavioral variables are based on the presumption that past performance is a true mirror of future performance. The success of RFM methodology (Recency, Frequency and Monetary Value) to predict future revenues from customers proves the efficiency of behavioral variables. The number and value of prepaid recharges made by a customer over a period of time, in addition to time lag from the last recharge can predict to considerable accuracy whether the customer is still with the service provider; it can also predict when and of what value will be his next recharge.
Segments are rarely based on one type of variable; a combination of multiple variables of different types can provide useful insights that define a customer segment. A study conducted by a colleague and me on the upcoming 3G market in India found that young (22-28 years) medium income (9–14 lacs) basic degree holders interested in music would have the highest willingness to pay for Mobile Gaming. Get entire report from here.
Once segments are identified and selected, the next step is to design service offerings and pricing keeping the target segments in mind. There is a trade off involved between the depth of segment definition and cost; some experts, however think that one need not make an either or choice. To read more about creating service offerings for a segment size of one, refer to the blog on 1to1 Marketing