Poindexter Systems President Joseph Zawadzki considers the advantages of pairing marketers with statisticians.
Over the past weekend, I was involved in a number of conversations that all centered around synergy between different disciplines (what E.O. Wilson calls Consilience) -- the concept that by simultaneously approaching a problem from two divergent points of view, you could create some novel solutions. This led to some intense speculation about online marketing, and eventually, to an idea or two about what might be in store for our industry in the coming years.
One conversation was with my father-in-law, who was in town to tour the facilities of a doctor leading his field in the development of cures for an inherited eye disease called retinitis pigmentosa. This particular doctor has made tremendous progress in his research with the Foundation Fighting Blindness by hiring a combination of PhDs in both medicine and electrical engineering. Instead of accepting the status quo and the current state of equipment they use to answer their medical research questions, they leveraged their combined knowledge to rigorously reformulate those questions and then build the specialized machines necessary to answer them.
This conversation, along with one about a hedge fund that put fixed income, debt and equity groups together to force daily interaction, inspired me to think of what might be a similar unlikely or unexpected pairing in the world of online marketing. I believe that the next real period of innovation in our industry will be realized through a "joining-at-the-hip" of marketers and statisticians. As we finally enter the world of branded response, where every touch point with a consumer is understood as a step toward a deeper relationship with the brand, and as increased spend requires increased accountability, we'll see a delightful pas de deux of solid marketing instinct buttressed by rigorous analytics taking hold. A handful of the things we should expect:
Counterintuitive insights
One of the first things that an advertiser must answer is "who is buying," with a special emphasis on those buying off the media plan (via tonnage advertising, search or passive site visits). For example, an apparel retailer discovers that those buying online are not their target upscale urban demo but a decidedly lower-income, more rural audience. Why? Because the higher-end market jumps into an SUV and drives to the store while others purchase directly from the site. The lesson is to beware judging the media as successful in reaching the former based on the purchases of the latter!
This is similar to the classic 7-11 study that found sales of beer increased by placing it next to the diapers ("Honey, pick up some diapers on the way home"), or to the experience of a lingerie retailer who found that their second largest target audience for stockings (after the expected 35- to 55-year-old women) were male owners of Harley-Davidsons who wear them underneath their leather pants to reduce chafing.
These lessons were learned simply through better data analysis free of existing prejudices or assumptions. By sitting a marketer alongside a statistician and asking them "what's unexpected and how can we do this better" we can arrive at some counterintuitive insights that will improve our effectiveness.
Consumers as marketers
Having read, and thoroughly enjoyed, Joseph Jaffe's Life After the 30-Second Spot, I see that a lot of smart people are rethinking the traditional relationship between brands and consumers. Jaffe talks about engagement, interaction and evangelism as some of the new ROIs that need to be considered as consumers are recruited to manage their own relationships with brands and in some cases even help build them. This makes a lot of sense in the age of TiVo, but again, both art and science must work together to craft a strategy through this next phase. Statisticians and marketers must ask questions like:
- What's important?
- How do we measure it?
- What does an extra second of engagement mean?
- How do we find influencers, and what is their influence worth?
- What's the net present value of an additional util of "community" (e.g., another entry on a blog or visit to user-generated content)?
It makes little sense to empower consumers by enlisting them as marketers in our own campaigns unless we know best how to harness their influence and how to measure it. Without this measurement and forethought, we are marketing blind -- rolling the dice and hoping that we do not create a monster in place of a powerful advocate.
Mass customization
"A Car for Every Purse and Purpose" began the trend toward mass customization, where marketers understood the individual needs of each segment of the public and a product was developed specifically to meet that need. It made perfect sense in the 1960s when Sloan unveiled the new slogan for GM, but the success of the strategy has sown the seeds of its own demise. Another current book, The Paradox of Choice: Why More is Less, illustrates how the great number of product options presented to consumers today has actually become paralyzing rather than liberating. With hundreds of automotive brands, toothpastes, and HDTV sizes and resolutions, any purchasing decision requires a tremendous amount of research.
There's a great psychological experiment that observed how people would respond when presented with a table full of jams and jellies in a supermarket along with a discount coupon if one of those products were purchased today. When the table had six products, 30 percent of consumers chose their favorite product and purchased it. When it contained 24 jars of jelly, less than three percent actually made a purchase. As evidenced, mass customization or too much choice has "jammed up" the decision-making works, leaving both buyer and seller worse off.
The marketer and the statistician, working together, could rightly determine when to offer the Model-T ("any color, as long as it's black") or the full panoply of choice for a given consumer based on data studies like the jelly conundrum. By better understanding the consumer, where they are in the lifecycle of purchasing a product category, and how they view a brand in that context, you can better decide what to market and when (if at all). For the consumer, this means the full range of options is narrowed down to those most likely to fit, reducing confusion and allowing them to experience something they might previously have overlooked. Again, a good quant-driven marketer will be able to learn what cues -- demographic, behavioral, modal, et cetera -- are relevant in a purchasing environment, and what to do in response to those cues to improve the bottomline.
Joseph Zawadzki is founder and president of Poindexter Systems.