Customer Data House-holding OR grouping of data
This is basic level of grouping (not clustering OR segmenting), which is applied with people having the same address OR/and same telephone number. example- You can add a common flag like 'family', 'Office', 'Department' etc. to enable you to send only one mailer instead of sending it to each of customers
Customer data Derivation on the basis of the data available
Example- You have the age of the customer, which you recorded at the time of the registration. Given the registration date you can do data derivation for year of the birth (Or a standard date of birth), so that you can avoid time decay.
Example- You have the Customer city given; you can fill up the state and the country. Same will apply on PIN codes.
Cluster Averages, means
Example- Let's say that you don’t have the income group for a customer. You can fill-in the blank by taking the average income of group of all the other customers having the same level of education, in a given city and the similar age.
TIP- Don't apply cluster and means approach if the reference population size is not large. Also if the blank data is larger proportion (say 10%+) of the population, one should avoid using averages.
Data Extrapolation and Data Interpolation
Example- If 5 years back the customer was 30 years and his income was USD 30000, you can apply a standard inflation and change his income accordingly. If at the age of 25 the customer was ' junior management' and had an MBA degree, you can place him at middle management at his current age of 35 years.
Using most probable value
Example- If the customer has an income above USD 25000, and he is above 30 years, the field 'whether taken mortgage' will have most probable value as 'yes'.
Field Survey
This of course is the most reliable, but the costly way of getting the information. Typically, this is done by third party companies in Database marketing OR large companies. Business case for this kind of exercise is generally provided for certain set of high potential OR high value customers. Through enticements OR as part of the business as usual interaction, one can get the additional/missing information from the customer. For example- Customer calls-up for a query and you can try to get his latest income status OR the number of dependents.
TIP- In case your business as usual interactions do not allow you to have the real data on your customers (or potential customers), we have tele-calling with right perspective helps enormously. This however works with your existing customers. If you are able to have the right positioning and communication scripts. There can be concerns about the privacy laws, and you need to focus on the information, which is important for your business. You may need to ask 'the type of company you are working with' instead of 'which company you are working with?' |
