High-Definition Data

Considering both the media scrutiny-as well as the political outcry-over data-collection processes employed by a number of high-profile U.S. companies, one might assume that when it comes to customer knowledge, American business knows it all. The sheer volume of data collected is massive; the number and variety of sources from which consumer information is gleaned-from credit applications and transactional histories to self-reported surveys and website registrations-appears to be endless. And did we mention the astonishing abundance of readily available public information?

So why, in this era of high-definition digital possibilities are a surprising number of otherwise sophisticated marketers still viewing their customer information in black & white?

For one thing, organizations are not seeing their customers in full view at all, but in isolated snapshots. That’s partly because the majority of legacy information management systems still in place throughout the business world can’t fully and efficiently integrate data from a variety of sources. In addition, many marketing managers are not sure how to accurately correlate information from multiple channels and draw conclusions they’re confident enough to stand behind. And finally, pre-2000 infrastructure simply wasn’t designed to support a 3D customer view. Consequently, like the blind men describing an elephant, managers find themselves making multimillion-dollar marketing decisions based on what they know, which is limited to only one part of the beast.

But those who embrace the new age of data profiling can gain new, revealing and potentially profitable insights about their customers and prospects. By appending data from a variety of sources to their customer files and utilizing advanced modeling tools, marketers are making relevant and actionable discoveries about their customers’ attitudes, orientations and ultimate behavior. And the resulting strategic plans are bearing fruit.

Consider these real-life examples:

  • A major financial institution, selling both retail and business products, was confident that its “sweet spot” among business customers was larger companies (i.e., those with $5 million or more in annual sales). In fact, not only did its marketing strategy focus on larger organizations, but its salespeople were actually encouraged to ignore smaller ones. Appending data on company size to their customer files and performing a strenuous regression analysis, however, revealed that more than three-quarters of its aggregate annual profit came from companies with sales of $1 million or less. “You screwed up my business plan,” said the company’s Chief Marketing Officer when he saw the results. “But better to know that now than find out three years from now.” The discovery triggered a completely new strategic focus on marketing to smaller firms.
  • A large direct marketer and retailer of desktop computer products and peripherals possessed data showing that its average high-end customer generated sales of $1,200. That figure became the cornerstone of its customer development strategy. But profiling analysis, built around consolidating all transactions eminating from a single family home address, showed otherwise: The actual value of that customer was $2,358, almost twice as much as the original assumption. Obviously, this finding led to a new, much more robust CRM strategy.
  • The publisher of a leading magazine for parents was challenged with turning around declining subscription renewals. Data appending, modeling and analysis revealed that its best candidates for renewal-and those most likely to renew at full price-were parents of children under 8 years old and, moreover, that the presence of both a boy and a girl predicted significantly higher renewal rates than those exhibited by parents of a single child. Armed with this information, the Circulation Director was not only able to tailor his renewal efforts far more precisely, but was also able to boost his new subscriber acquisitions efforts by targeting parents of more than one child.
  • A leading distance-learning provider observed that a relatively small percentage of those who enrolled in its programs actually completed the full curriculum. Those who did were very profitable; those who dropped out early were extremely costly. The challenge: How could it identify in advance which students were most likely to remain enrolled or drop out? Appending age and various lifestage attributes to its enrollment files and modeling against student payment history provided the answer. The company is now able to target promotions to market segments most likely to stay enrolled, dramatically reducing its attrition rates.

Profiling tells you what you don’t know

The core of any profiling program is an organization’s existing house file, complete with its transactional data. The date of the customer’s initial transaction, what item or items were purchased and through what channel or channels, the profitability of the transaction (with a solid understanding by all parties on how profit is measured) and any follow-on sales are all taken into account. The goal is to identify your best customers, so that you’re in the best possible position to go out and find others who “look” just like them. Regarding your less profitable customers, while your organization may not want, or be able to “fire” them, knowing this information can lead to a refinement in strategy that will result in a closing of the gap. Once you know who your customers are, and how much each is worth to you, you can make informed decisions about how much to spend against each of the various market segments you serve.

Recency of transactions is widely recognized as a main driver in many companies’ ongoing selling efforts, and it’s an equally critical factor in profile analysis. Annual attrition on a typical catalog or retail file, a magazine subscriber file or a non-profit organization’s roster of donors can be 50% or more. It would be a miscalculation to assume that, just because the fundamental purpose of your product or service hasn’t changed your buyer’s motivations or characteristics also have remained the same. Customer databases are dynamic and must be analyzed regularly to make sure that basic assumptions supporting your business model are still valid.

How do you determine which attributes to key on in a profile analysis? Most experts agree that it’s probably equal parts scientific examination, intuition and maybe even a little magic. Certainly, it’s a great deal of trial and error. In any case, it’s crucial to keep an open mind and maintain solid analytical capabilities. The financial services marketer cited above was sure small businesses were not his organization’s prime target, but was willing to retool his strategy once it was clearly demonstrated otherwise. The publisher still can’t quite understand why two children-in particular a boy and a girl-should be more viable prospective subscribers than a family with a single child or two girls or two boys. But both were prepared to capitalize on the discovery of the unique data “spike” that signaled a defining difference between their optimum target markets and the general population.

Donn Rappaport is the Chairman/CEO of American List Counsel, Princeton, NJ. This article is re-printed with permission from The Rest of the Story, published by ALC, a data and marketing services provider dedicated to enabling their clients to grow, increase market share, and improve profitability through the innovative gathering, enhancement, integration and strategic use of data to improve customer acquisition and customer value. To learn more about The Rest of the Story or to subscribe, please go to

Comments are closed.