Gain insight into your customers

Profiling provides a means of understanding key characteristics that define a group of customers. Profiling can be undertaken on a demographic basis or on a behavioural basis, or a combination of both.

Demographic Profiling

Demographic profiling "describes" a segment, for example, age, gender, location, business type, organisation size or job function. Demographic profiling can be used to identify prospects who share similar characteristics to known customers or to determine suitable brand values, where to target advertising or how to promote suitable website content.

Behavioural Profiling

Behavioural profiling "describes" the behaviour of a customer segment, for example frequency of purchase, product take-up, website visits. Behavioural profiling should be used to determine how a customer is interacting with your business, how to promote relevant cross-sell offers or what level of discount should be offered.

Which One?

Behavioural profiling is more predictive than Demographic profiling as it gives real insight into actual customer behaviour. Perhaps the best example of this would be the RFM segmentation method which uses previous purchasing behaviour to classify customers.

It should be noted that profiles change so they need to regularly updated. Lastly use propensity modelling to test the significance of any demographic profile value on your customer database as the existence of such a profile value on a record is not necessarily a significant factor in the customer's purchasing behaviour.

A sample profile report and what it means

1. Profiling compares one group of records with another to identify the characteristics that differentiate them.

2. Profiling detects characteristics by identifying results that occur in your selected data and which are highly unlikely to occur in a truly random sample of date.

3. Profiling is typically used to find the characteristics of a group of records.

4. Profiling one group of records as a subset of another is a penetration profile.

5. Profiling one group against another separate group is a comparison profile, for example identifying similar or differential characteristics between a test and control mailing.

6. You can determine which variables to include in the profile, and what level of detail is included. You may leave out variables because of what you know about the meaning of the data and how it was collected.

7. Profiling allows the user to identify attributes significant to the next stage of modelling.