Avoid Information Overload
Look to your ultimate business objectives to determine what data is worthy to collect and keep
By Irene Cherkassky
Common practice suggests that when you start any endeavor, you always should start at the beginning. However, when it comes to data collection, it’s far more advantageous to start at the end.
“Trying to decide what data you want without having a very clear picture of what it is you’re trying to do—what are the business objectives—[is like] shooting in the dark,” explains Chris Lucas, vice president of product management for the Sales and Marketing Solutions group at Short Hills, N.J.-based D&B, a global provider of B-to-B data information and solutions. “It starts with a clear articulation of what it is [you’re] trying to do, then work backward to the data information you need to support that business’s objective.”
Having a clear understanding of your business’ goals will help you home in on data elements that are most relevant, leading the way to a database that can underpin effective marketing campaigns, boost profits and be cost-effective.
One of the most important questions that will help shape your data collection is whether the data will be used primarily for customer acquisition or to increase the value of your current customers and manage communications with them. The answer will help guide your database design.
For instance, on the consumer side, Maria Marsala-Herlihy, senior vice president, strategic consulting and analytics for KnowledgeBase, describes, “There are probably 10 to 12 basic demographic fields that are going to be helpful for everybody, but are critical on the acquisition side.” This key suite of elements includes age, income, date of birth, net worth, presence of children, marital status, homeowner status, what type of house they live in, family composition, and length of residence. Additional behavior-based factors to consider include:
* How many mailings did it take to get a customer to convert? This will help gauge what it will take to acquire more customers.
* If more than one creative is used to gain new customers, which was more effective?
* What time of year did the customer convert? Seasonality is key to understanding where and when a product or service will be more in demand.
If the focus of your database purely is managing your own customers, rather than acquisition, you must ask which pieces of information are going to significantly impact your ability to increase the value of a customer, says Marsala-Herlihy. Recency, frequency, monetary value (RFM) data are essential to any continuing marketing efforts, she says. As well, understanding the details of the customer’s first interaction with your company is vital. “[In] every lifetime value study I’ve ever done, that first interaction is so telling,” says Marsala-Herlihy. “It’s very predictive long-term, and that’s almost across all industries.” This includes what customers bought, how much they spent, whether they bought across categories, and how many communications it took for that customer to convert.
Much of your decision as to what data are relevant to you will depend on the type of marketer you are. For instance, healthcare marketers, says Marsala-Herlihy, are likely to be more interested in acquisition data, since they are subject to severe privacy restrictions as it relates to using customer information for anything other than what’s necessary to provide the subscribed healthcare services. In comparison, most retailers will be more focused on transactional elements.
B-to-B marketers need to design their databases using similar building blocks. According to D&B’s Lucas, the seven elements essential to any B-to-B database are customer contact information, business name, address, telephone, industry SIC code, number of employees and sales volume. “Many businesses use these [elements] exclusively to do segmentation and targeting,” he points out.
Making Sense of the Data
Once you have collected the relevant elements that are the foundation of your database, companies such as D&B and Little Rock, Ark.-based, customer and information management solutions provider Acxiom Corp. offer the tools to help marketers get the most out of their data. Acxiom, for example, offers data portrait analysis using InfoBase, its compiled consumer database.
“If you already have a population of customers … what we do with a portrait analysis is take that household data and match it against the 200-plus variables on InfoBase and look at the core characteristics of those [customers],” says Josh Herman, InfoBase product marketing leader for Acxiom. The next step, says Herman, is predictive modeling, which helps to identify the elements that are predictive of purchase behavior—again, this includes elements such as recency or purchase, age of head of household, and income.
D&B offers tools such as its Integration Manager solution, which global enterprise software provider Oracle recently has implemented to get a more up-to-date, complete and actionable view of its customers. Using Integration Manager in combination with its own E-Business Suite and Customer Data Hub, Oracle can better leverage its data to create more effective sales and marketing campaigns. Oracle’s customer and prospect information data records are automatically matched against D&B’s global B-to-B database and is assigned a D-U-N-S Number (Data Universal Numbering System), a unique nine-digit code that helps identify and link more than 100 million companies worldwide. As well, those Oracle records also are appended with any additional information from the D&B database, including corporate hierarchies, predictive indicators, and demographic information, such as number of employees, revenue and age of the business.
According to Oracle, the use of this process has resulted in as much as a 300 percent to 400 percent increase in ability to accurately perform market segmentation and a 30 percent reduction in duplicate records, thanks to the
D-U-N-S Number, which helps identify redundant records. Also, as a result of the matching process, Oracle now can see corporate family tree relationships.
“Through D&B Corporate Linkage, we’re able to relate multinational locations together, and that’s a real advantage for us,” says Thomas Brauch, Oracle’s director of analytics, global marketing. This helps identify global opportunities. Brauch notes, “D&B’s marketing intelligence helps us determine which companies to contact, with which product, and at what price.”
Keep in mind also that collecting data and building a database are two different things. Various elements may be collected, but not all of them are going to be valuable for the purposes of your database. “It’s a two-step thing. One is collection of [the data] and keeping it offline and on tapes, and the other thing is determining what you actually need to keep online and accessible,” Marsala-Herlihy says. “Put in the database fields that are drivers of the consumer behaviors that can make a difference in revenue or profitability. Store everything else you are collecting offline.”
When it comes to data retention, Marsala-Herlihy offers the following guidelines: “Keep online and available two to three years’ worth of data, and then I usually recommend keeping an additional two years offline, but available for up to five years.”
Lucas adds that current technology offers marketers alternatives to collecting data files in bulk and maintaining them. “With advances in technology and standard Web services offerings, advances in customer data integration actually allows you to pull together information that’s not stored together quickly and efficiently.” Now, says Lucas, customers can pull the attributes they need from individual source systems on demand.
Is It Worth It?
Ultimately, data collection should be viewed like any other business investment. When deciding what data is worth collecting, Mary Ann Kleinfelter, director of marketing, direct mail, for Baltimore-based tutoring services provider Sylvan Learning Center, says, “The real acid test is whether this data is something the marketer can use to benefit customers—which will increase your sales and profits.” Therefore, as with any other investment, you need to figure out the cost of the data to you, which includes the cost of acquiring the data plus the cost of keeping it clean and updated. To calculate whether that cost is worth it, next you need to find out how much money that data is going to help you make, says Kleinfelter. Testing is the best way to calculate that worth. For example, if you’re trying to determine whether a factor such as high income or low income will influence your campaign’s outcome, begin by mailing an offer to customers you always mail to as your control. Then mail a separate group of high-income people. “Based on the lift you get from the high income, or the money you save by not mailing to the low income, that’s how you can quantify what that income selection would be worth to you.”
“It comes down to a mathematical equation that says: Here’s how many sales and [how much] profit I got when I didn’t use the data, and here’s how much I got when I did use the data. Is that amount of money going to be greater than what it’s going to [cost] me to buy that data and clean it?” explains Kleinfelter. “It’s really simple, but you’d be surprised. A lot of people don’t do it that way.”
What Not to Do
Whatever type of marketer you are, the least advisable way to decide what is relevant and worth keeping is by making a guess based on your intuition, says Kleinfelter. “What a lot of people do is say, ‘Well, I just know I’m better off with higher income people,’” she describes. “They just rush out and buy high-income information and they don’t measure to see if it gave them the kind of lift to pay back what they are spending to get it.” If you’re basing data collection and upkeep on hunches, warns Kleinfelter, your ROI will suffer.