One of the more frequent questions from marketers I hear is “what (or how) attributes (data fields) can we do segmentation on for our campaigns”. The answer is usually more or less this; almost all data in your CRM system can be available for use in the key process in many CRM software systems. Let’s take a look at the data for segmentation. Some of the data is straight forward and relatively easy to understand and acquire, but other data may take some deep thought and effort to produce.
Major Attribute Categories
In most CRM systems, they will have a segmentation process that involves the ability to select contacts that will create a “list” which in turn will be used when a campaign is executed. This list contain the necessary contact information to assign that contact to the campaign. Campaign segmentation data categories are:
-
Base contact information
- Name and address information
-
Extended contact information
- Phone, email addresses, billing addresses, local and regional metropolitan areas, etc.
-
Demographic data
- Information about where this contact fits into many different areas
- Gender, income, ethnicity, home ownership, mobility, locality mapping, age, etc.
-
Psychographic data
- Subjective information about the contact or their subjective responses. These are abstract characteristics related to ideas, opinions and interests.
- Examples: I like to play tennis, I go to restaurants 5 times a week, I like cats and dogs.
-
Transactional data
- Information that is collected as a result of the contact’s interaction with the company.
- Examples: Order and billing information, Products purchased, Web usage, eMail usage, call center interactions
-
Response Data
- Information that is collected into the CRM system as a result of a previous campaign interaction.
- Examples: Contact clicked on the promotion eMail, they declined the offer, they purchased based on a promotion.
-
Segmentation Assignments/Score/Indexes
- Segmentation processes can be design where they run against the contact population and using segmentation filters from the data above, create new values, indexes or scores that are recorded into the CRM system. These new values then can be utilized like any other attribute in creating a new segment of contacts for a campaign.
- A typical segment score would be a financial score such as FICO where all the information about your finances, location, etc are put into a calculation routine and a new score is developed that in a single value ranks your financial “power” against all other ranked contacts.
-
Predictive Models
- Using complex algorithms marketers can develop statistical models that use existing contact data and attempt to predict what the consumer will do in the future.
- An example of this might be to predict the likelihood to purchase a selected product during a given timeframe.
All of this information is fair game to the segment process. Choose values in any of the data attributes in any combination and you can have a segmented list for the next campaign. One need by the marketer is the need to go beyond the above categories.
The random segmentation process
Even when one has a great amount of data about the contacts, marketers want to use non- attribute based methods to create a segment list. This might be used to create testable splits of a contact universe, A / B split testing. Of course this actually can be split into N number of segment list if desired rather than the two as in A / B tests. Generally your segmentation process will consist of initially creating the universe that you will apply the random process to. Sometimes it might be your entire contact universe or it may be some selected sub-set of contacts. In either case, once the desired universe is selected a random process of choosing the contacts is initiated. For the random process, there are many different methods but generally the idea is to create a split of the available universe into the A / B segments on a completely random basis. This allows the marketer to not influence which contacts get into any of the segment lists.
Plan your attribute data and segment processes
The use of segments in marketing is key to the success of a company’s marketing efforts. Segmented lists provide the ability to target and create segment lists that make sense. Thus when designing any campaign or creating the overall CRM system, think about what segment data you would like to use.