The last several years have brought much discussion about techniques like customer relationship management. Some campaigns have worked spectacularly; others have not fared as well. A good explanation for the less successful efforts may be a poor grasp of customer data.
It’s important to remember throughout the consideration processes for advanced technologies that such efforts should be driven by a solid understanding of the basics. Without this, most projects will be doomed to fail.
Now might be a good time to review and refine the foundations of your marketing and customer systems and to evaluate the factors that affected the success of your more lofty customer-related initiatives.
Data, Data, Data
The never-ending battle, as always, is handling data, the lifeblood of our systems. The challenge is like the motto for the board game Othello (“A minute to learn, a lifetime to master”). Likewise, a simple and well-planned relational data scheme can accommodate not only existing data needs, but also can function as the basis for growth of data collection processes and support additional functionality.
When properly implemented, many information systems can help to keep straight the volumes of customer data available in a large organization. Without proper planning, though, errors and inconsistencies can creep back into the information repository.
Getting back to basics with data is simple:
- Clean your data
Make sure your data sets are clean and that all fields are as complete as possible.
- Examine your data rules
Double-check data rules to make sure the existing data complies with stated field variables and requirements.
- Review the utility of your data
Examine your existing data usage and functionality to make recommendations about any new elements you might need.
- Look for opportunities to append
Learn more about your customers by appending additional data such as demographic, firmographic or psychographic elements.
Strategic Mindsets
Proper use of data can provide a tremendous competitive advantage. Anything that is not done wrong (for example, sending acquisition messages or duplicate messages to established customers, typographical errors in name and address, etc.) can create loyalty, just as easily in that doing things wrong can breed disloyalty or cause customers to move to your competition.
Do you have strategies in place to keep data current? Do you contact your customers periodically to confirm names or e-mail addresses? Do you use Web and point-of-sale scripting to fill in missing preferences and phone numbers, or do you occasionally conduct small surveys to get an idea of behavioral trends and satisfaction with your own performance?
The result of effective strategic planning will incorporate and support growth in customer loyalty and strengthen your customer relationships.
Consider the basics of the strategies used to manage your data warehouse as well as what you can do to leverage routine data usage and maintenance:
- Data stewardship
Often a small team of people in the organization handles data hygiene and utility. Note who’s responsible for data rules and data manipulation. As needs change, data marts and odd data sets may clutter up your systems. Consider those accountable for tracking data and who is charged with keeping a centralized data dictionary up to date.
- Multichannel communications
Each different channel and interaction provides a chance to collect incremental information. Consider your market plan from the standpoint of what additional knowledge can be acquired as a result of such interactions. Also look at how each channel is being used to communicate messages. This is a chance to ensure that you’re not sending conflicting messages to your customer base.
- Evolving strategies
Plan for the unexpected. When establishing a yearly marketing plan or strategy, you need to take variations into account. Always intend to have a small reserve budget specifically for new tests (format, audience, offer, channel) in order to fine-tune and optimize your channel use, messaging or simply to accommodate ad hoc communications.
Analyze This…
Of course, the basis for all this is sound and methodical analysis. In order to clean and optimize data, you’ll need to perform the basic steps of exploratory data analysis (EDA) such as running field frequencies and checking for missing values, exceptions to data rules and the like. EDA really should be thought of as an iterative procedure rather than a one-off snapshot process. Also, by having a good grasp on the state of your data and existing routine analyses, you can become more proactive in handling and accommodating changes to your needs and system.
While tools exist to perform sophisticated data analysis in an automated environment, much of the basic work can be conducted with desktop tools such as Microsoft Excel and Microsoft Access (pending the size of the data set being examined).
Beyond reviewing data coverage and quality (not capturing necessary data elements, campaign codes, dates, gender, phone numbers, e-mail addresses), the next steps should include fundamental segmentation and creating customer and product profiles. This will help dictate what additional data fields might be required as well as direct the development of strategies to collect or manipulate information.
Before running sophisticated models, you can discover a tremendous amount of information that will either confirm or contradict your premises about your customers. These analyses include looking at trends such as changes in segment sizes, RFM redistribution and looking for new or emerging segments. These efforts also may actually influence the direction of, or help build a case for investing in, more complex CHAID or regression models.
From a customer standpoint you can evaluate the strategy behind maintaining or updating profiles that periodically refresh the picture of your customers (age, gender, geography, income, spending patterns). Purchase patterns, such as multichannel buying behavior, can vary over time and cause changes in strategy or communications.
More detailed studies may include integrating behavioral information with motivational perceptions, such as primary research, surveys and panels. Regardless of how sophisticated your current analytical capabilities happen to be, these steps are critical to driving and adjusting the cycle of customer interactions.
Network
