Customer and prospect data is the fuel that runs your Customer Relationship Management (CRM) and direct marketing campaigns, and the internet has enabled marketers to gather far more data than ever before. The problem is that most of that data currently gets wasted.
I suppress a groan every time I hear a client or vendor crow about all the customer data they have collected while showing me a long, unstructured list of web page visits, promotions sent, and even products bought. 9 times out of 10 they cannot tell me what all the files mean or how they are going to use that data. This represents a lost opportunity. You have to plan ahead and know what you want to do with customer related data so that you can classify and structure the data to best meet those needs.
The goal of a CRM program is to enable your database to replicate what is intuitively obvious to a human. When working with a client, I encourage marketers to picture themselves as watching a person explore their store or office and imagine each page visited, or email responded to as a customer picking up a product or brochure. For example, let's say one prospect came into a travel store, stopped to look at brochures for a cruise, resort, and airfare for the Caribbean. The prospect took information about economy trips and discounts whenever possible. In another case, a prospect looked at a brochure for a cruise in the Caribbean, a cruise in Hawaii and a cruise in the Mediterranean. The second prospect took information about how to upgrade to first class cabins whenever possible.
If you were a sales person sitting in that store and had the ability to send a follow-up direct mail piece to each customer, you would have a good idea of what you wanted to send. You would probably send a discount Caribbean focused promotion to the first and a luxury cruise-focused promotion to the second (preferably in tropical climates) to the other. If you structure your data correctly, your database can make the same conclusion you would if you were able to watch each interaction of each of your customers/ prospects.
How to Structure Data Correctly
The first step in enabling your database to make the same conclusions you would is to create a category structure for every communication with customers/prospects. The goal of the category structure should be to enable the marketer to speak to each customer as an individual. The structure should be focused on the interests of the customer and should also apply to your products, content, and promotions. Note that the customer focus of the categories of interest usually prevents using the existing structure of product SKUs, which are grouped by brand, product group, or structured specifically for inventory management. However, the interest categories should be mapped to your SKU database or any other pre-existing product or transactional database.
If we look at the travel example above, the categories created in the database will be destinations, travel types, and lodging types. Locations can be as extensive as a list of every possible destination, whether they are US cities, or third world resorts; or areas, such as the Caribbean that would cover all locations within the Caribbean. I would recommend not making the initial categorization structure more complicated than necessary. Start with a simple categorization structure and then expand the detail of your categories as you learn about your customers with each campaign.
Groups and Relationships
Once the initial geographic categorization is created, it is important to create category groups and relationships. For example, The Caribbean and Hawaii could both be classified under vacation destinations, islands, and tropical destinations so that every time either Hawaii or The Caribbean is chosen, it is automatically recorded as each of the three classifications. This allows the database to draw similar conclusions that a sales person would. For example, if someone looked at information about the Caribbean, Aruba, and Cancun, The database would show an interest in tropical vacation destinations in the western hemisphere if they were all classified as members of those groups.
Creating category groups can be especially useful when trying to determine the interest in different types of products or communications. In the given example, two potential groups would be Travel & Lodging types. Let's say Travel & Lodging types consisted of Air, Train, Cruise, Resort, Camping, Hotel, and RV. If you simply made these seven Travel & Lodging types separate categories, you would not be able to tell how they varied by geographic region or destination type.
For example, one person might prefer 'Camping' in the 'Midwest', 'Resorts' in Rockies and 'Cruises' in 'Tropical Areas'. Without directly linking the Travel & Lodging type directly to the destination, you would be limited to the knowledge that the person has some level of interest in each type. As a result, you might send a promotion for a 'Resort" in a 'Tropical Area' that would not interest the person. In order to target by both type and location, you have to create subcategories for each Travel & Lodging type under each location.
It is then important to group each of these types together so you can still draw aggregate conclusions per individual or region. In other words, every item categorized as a cruise, regardless of the destination will add the individual's cruise preference. Only then can you effectively determine each person's interest on both an aggregate level and for each location. As mentioned it is best to evolve to this level of complexity after starting with a simple and robust category structure.
Implementing a data categorization structure provides insight on aggregate customer trends, customer segments, and enables marketers to treat each customer/prospect as an individual. On an aggregate level, you will be able to determine the correlation between different types of categories. For example, you might learn that high-end vacations are popular in Hawaii and low-end vacations are popular in the Caribbean. You might find that free upgrades are most effective for cruises and that discounts are most effective for airfare.
On an individual level, a well designed data categorization structure will allow you to modify your communications for customers and prospects that you do not know enough about to personalize based on his or her individual interests. It also allows you to better merchandise offers on your website. Perhaps more importantly, you can customize your communications to each customer based on his or her trending interests.
Categorization is only the first step. Next time I will discuss how best to drive individual communications based on the categorization structure.
Please feel free to ask questions of or provide feedback to the author at pmeyer@ifinder.com
