Finding consumers and figuring out how to market goods to them is a task that's becoming more and more complicated these days. Typical marketers use four or more data sources to profile consumers. Then, they use 10 or more media measurement sources to figure out how to communicate to their targets. Analysis and execution become messy and generalized when you try to coordinate information across these more than a dozen independent sources with assorted, non-unified measurements.
Good consumer profiling demands a variety of information, from purchasing data to advanced target area models and loyalty vendors. But all these sources don't report much on consumer attitudes, usage patterns, and motivations. A second layer of research often investigates these factors with custom, product-specific attitudes and usage (a&u) studies. After distilling what product attributes and consumer attitudes drive sales and what the competitive landscape looks like, marketers usually investigate key characteristics further in focus groups of like consumers.
Serious money is spent by marketers to unearth core insights, yet at the end of the day, most of these insights are trapped by dead-end data. The fact of the matter is that product-specific attitudes, usage, and key characteristics are not measured by the 10 media measurement sources available. So how do you execute your marketing target? The industry answers include using the imaginary single source, mimicking the target with surrogates, or investing in data integration and management.
Personally, I do not understand the single-source model. There is too much to explore about who consumers are and why they behave in a certain manner to think some generic data collection would work for all products and all media. This is the horizontal versus vertical issue: Horizontal researchers collect a wide variety of information, and thus don't investigate anything too deeply.
Surrogate targets are the common solution that link "who" studies to the many execution-based media studies. This "mimic method" distills marketing targets into simple "media targets." Often, the simple targets are age, gender, and income, as media "currency" studies lack detailed information on attitudes, usage, and other differentiating factors that constitute good marketing targets.
This is sometimes disguised with intermediary horizontal studies that measure general attitudes, usage, and media consumption. In these cases, in-depth marketing analyses are converted to intermediary-speak to provide media understanding of the "marketing target." This approach assumes that horizontal and vertical marketing and media data are identical, or close enough.
Weighing vertical media decisions with horizontal direction may seem to work. Yet most practitioners do not know about the significant differences between horizontal recall data and the electronically measured vertical data, because of their traditional age/gender perspective. Dig into geo-demography and significant differences crop up in product and media data; they're not fuzzy and not good.
The strategic purpose of the horizontal studies is not the misplaced notion of single source, but the elegant glue of data integration. They are perfectly placed to link vertical studies in purchasing, product attitudes, and media consumption, while providing marketers with information beyond category insights.
There are two basic schools of data integration: fusion and segmentation. Fusion links specific respondents between two studies, while segmentation links groups of respondents between multiple studies. Personally, I think fusion is a waste of time. Who wants to mono-focus on two studies?
The Internet is tipping research into database marketing. As marketers increasingly use nonrepresentative Internet research to cut costs and gather quick information, they are awakening to the potential of crossing the line from pure investigative research to building relationships.
Mark Green is senior vice president, media services, VNU Global Modeling & Analytics, and the founding partner of the Media Learning Institute. (email@example.com)