For one-to-one relationship marketing to be successful over time, it is paramount that the relationship effort be built using sound methodologies. Following are some steps crucial to creating a turnkey, database-driven, one-to-one relationship program.
Keep in mind that those numerous piles of data can provide valuable information on your current and potential customer base. Although a daunting project at first glance, building a powerful, meaningful database can be your ticket to a more successful, economical and meaningful marketing program.
The key to building a useful database and extracting valuable information is to follow well-thought out steps.
Prior to doing a model, it is necessary to fully understand existing data on the customer database. Data elements such as frequency of purchase, amount of purchase and other transactional data are key in understanding the characteristics of your customers and, more importantly, defining the characteristics of your "best customers". Understanding your best customers allows for mirroring their characteristics in prospect databases.
In addition to data currently on the customer database, appending business data from vendors such as Dun & Bradstreet and/or InfoUSA can also be used when building a model. Vendors that specialize in gathering and maintaining business data can provide useful data on businesses such as employee size, revenues or SIC categorizations. Match rates (the ability to match a customer/business on the customer database to the vendor database) vary but usually fall in the range of 40% - 60%. Typically, appending data usually takes two business weeks; timing varies by vendor. Prices for appending data are on a cost per thousand basis, usually under $100/M, pricing varies by project.
Once data appending is complete, the modeler can begin to build a model. Because there are many types of models that can be built (Chaid, regression, neural, etc.) the modeler can assist in determining the best approach. Modeling typically takes a month but depends on the scope of the project. Pricing can be anywhere from free to $50,000 per model. In cases where the modeling is done at no cost, a vendor will typically require that appending of data be done at their facility.
Once the model is build, the modeler will usually create a report highlighting the following: the salient characteristics of the customer base, qualities of pre-defined "best customers", docile groupings of the customer base from best to worst, etc. The model can then be used to "score" or find similar "best customers" in a prospect database. Model reports vary by vendor.
Based on the results of the model, the characteristics that define your customer base can assist in segmenting them into specific groups or niches. Having segmented the database, unique marketing messages can be specifically created to "speak" or relate to each group in a unique manner. In addition, depending on the characteristics of each segment, communication vehicles may vary. For example, one group may respond better to trade journal print ads versus direct mail. Again, a commitment to continual testing over time is essential for the success of a one-to-one relationship effort.
The list (customer base or list of prospects), offer (what is being offered) and creative (the design, look and feel of the offer vehicle) will continually need to be tested over time. By changing the list, offer and creative over time, measuring response will provide insight into how to continually improve response.
To accurately assess how well a particular campaign is performing, measurement benchmarks and systems must be put into place. For example, a tracking code should be put on a direct mail piece to link the response to the record on the database so response can be tracked. Measurement can then be used to establish control groups and continually track response over time. Hence, the one-to-one relationship program can continually be tested and improved upon over time.
In conclusion, it is essential that a one-to-one relationship program be structured based on sound research and analysis. The analysis will then drive segmentation that can then be used to determine the target, offer and creative, hence creating a full-circle, turnkey approach to a database-driven, one-to-one relationship effort.
What We Do |
Association Laboratory Experience |
Press Releases |