Analytical Paradox

The business of customer analytics is to get answers quickly so marketers can leverage newly gleaned information for competitive advantage. Why, then, is it so difficult to get answers to simple questions, let alone the complex ones? Often, marketers are faced with a management report that, instead of answering questions, spawns a thousand different questions. These in turn create requests for new reports or data. And a new cycle of system development begins. However, therein lies the rub: Traditional analytic system development must have requirements, but to define requirements, marketing must have a system to understand their customers. This catch-22 phenomenon is known as the Analytic Paradox: "I can't tell you what I want until after I know it." Architecture of Strategic Questions For years marketing has been the "hard to pin down" discipline with respect to requirements. Unlike other business units, marketing's objectives are to change the status quo by being creative. This strategic focus creates demands on analysis applications unlike any other business applications. Another way of thinking about this: Strategic insight comes from finding something new and different in the way the market works or customers interact with your company. By definition, it is impossible to define what this is before it is found. Developing creative strategies is about finding and leveraging things that "I didn't know I needed to know." These are the insights that have the biggest impact on marketing to customers, whether for acquisition, retention, cross-selling, or weaning. Strategic questions also differ from the traditional tactical or operational questions in terms of their longevity. Reports such as sales by territory, in-bound calls by type and duration, inventory status, etc., are repetitive metrics that the business needs on a continuing basis. These metrics provide the ability to assess business performance relative to plan and budget. For this reason the "I know I need to know" questions are very tactical in nature and can be defined through a deductive process. Answers to strategic questions often have a very short half-life. The answer may be of such importance that the business focuses on it and thus becomes tactical; or, the answer has no value so it isn't asked again. So how do these one-off and strategic questions get answered? Traditionally, it has been a matter of fitting a square peg into a round hole. Using tools and technologies for management reporting and tactical intelligence to address and explore strategic issues often result in severe penalties (time, cost, or reputation) for asking things like, "How have our customers changed?" What is needed is a new approach to answering these spur-of-the-moment questions so marketers can quickly act upon the answers. There should be no penalties for asking strategic questions. More important, marketers should not apologize for asking unplanned questions. Architecture of Technology To deliver this information in a timely fashion, great pains are made in tuning the database, streamlining the report distribution process, and instilling rigorous structure and hierarchy in the information. This works well for tactical or management reporting, but this approach is ineffective (and in fact completely opposite) when feeding the insatiable appetite for new insight. The technology should be indifferent to what you ask and where you want to go. It should be just as adept at manipulating the non-buyer as it is the buyer, such as creating a segment of people who haven't bought in the past six months. There should be no proscribed paths for navigation. And all of this must perform at train-of-thought speeds so that the "Aha!" moment isn't lost while watching the hourglass, or worse, waiting for a new report or database to be built. Tools that provide this kind of flexibility generally don't come from the "row" world of transaction processing, such as those traditionally used for analyzing data from reservation systems, Web visits, or catalog orders. And it doesn't come from the "cubes" of management reporting, such as those used for financial budgeting and allocation, executive information systems or sales analysis. Whether this new class of tools are column-based analytic technology (CBAT), vector-based, or other techniques, they must structure information for querying large data sets. That is what enables users to perform "train-of-thought" analysis against the complete-detail customer and transaction histories. No more summaries that aren't quite right, or funnels that filter out valuable information simply because you didn't know it was valuable. Such a tool has helped Carlson Marketing, the world's leading relationship management company, provide more tailored, cost-effective services to its clients. It has helped reduce data preparation costs by 40 percent, which gives Carlson four more days out of 10 to provide additional services to its clients. Carlson also got a payback on its investment in fewer than six months. This was accomplished purely through increased efficiency, regardless of the income generated from additional business won as a result of having the technology. To survive in this more complex, dynamic environment, marketers need to get answers quickly to the simple questions. The Analytic Paradox is something best solved by using tools that can "converse with data" on your terms and deliver answers that don't penalize you for asking new questions.