Predictive skills require a willingness to weed out entrenched cognitive biases and established decision-making processes built into corporate culture
Marketing is in the midst of a precision crisis. The new world facing marketers has been described in many ways, each accurate in its own way, but none speaks to the fundamental shift now roiling marketing. Pundits preach second-order consequences of a deeper transformation at the very foundation of marketing. That transformation is one of precision.
The most over-used quote in marketing is the variously attributed statement to the effect that half of all marketing or advertising is wasted, we just don’t know which half. It’s a tired, old saw but it captures the very essence of marketing. This is the existential kernel of marketing, or the basic reality that has shaped how marketing is conceptualized, designed and executed. Simply put, marketing has been built on a presumption of imprecision.
Perhaps nothing illustrates this presumption better than the media model introduced in 1961 by the Advertising Research Foundation, developed to sort and organize various metrics for evaluating ad buys. It was depicted graphically as a left-to-right winnowing down of an audience, not the upside-down pyramid of consumer decision-making that it was turned into later. Either way, it represented marketing as a process of management through imprecision.
The basic idea behind any purchase funnel or consumer journey map is that marketers don’t know precisely who will respond—or even how—but if marketing can get in front of a large-enough group of consumers, sufficient numbers will make a purchase. Factors such as reach become critical in navigating this imprecision. For example, if past experience or research establishes that 10% of consumers reached will buy, then marketing has to reach enough consumers such that 10% adds up to sales or profit targets. In this way, imprecision can be turned into success.
One critical nuance is that the historic imprecision at the heart of marketing is that of individual consumers, not aggregate groups of consumers. Marketers have known with precision that, say, 10% of a group will buy if reached. What marketers have not known is precisely what individual consumers will be part of the 10% who buy. But it is the lack of individual precision that creates the cost and time inefficiencies.
Certainly, marketers have been looking for ways of operating that are better than intentionally over-spending to ensure a critical mass. But precision has been elusive and the traditional infrastructure of big agencies, big media and big budgets built to cope with imprecision has endured.
In the past decade, a sea change in precision has thrown marketing into crisis. Data, digital, algorithms and AI have unlocked the ability to study and engage with individual consumers in personalized and predictive ways. It’s not a revolution in data or digital per se, nor merely a handoff of execution to algorithms and AI. All of these things are critical and all play a part. But what’s fundamentally different is the precision at which data, digital, algorithms and AI now enable marketers to operate.
Having access to greater precision doesn’t automatically lead to better marketing. Lots of modeling continues to show superior returns for TV advertising over digitally directed ads. But these findings are misleading. The biggest impact of precision is less in placement and more in personalization. Getting in front of the most-likely buyers, and only those buyers, is most effective when it is done to deliver products and services customized for individuals. Predictive personalization is the power unleashed by precision.
Bad predictions create sizable inefficiencies that people must pay for and endure. No one wants to waste money on a product they won’t like or use. Better prediction can eliminate, or at least minimize, these sorts of unproductive inefficiencies. There is a lot to be said for serendipity and discovery, but not when they create improvident inefficiencies. Fewer inefficiencies will have a dramatic impact on people’s lives.
Many marketers might feel that predictive personalization is old news, but such prediction is in its infancy. Entire categories are on the verge of being wholly upended by predictive personalization, including foods, beverages, financial services, media, entertainment, retail, travel and insurance.
In a recent interview with Strategy+Business magazine, Piyush Gupta, CEO of DBS Bank of Singapore, explained that the business model of insurance is one of “socialization of risk,” which is to say it is rooted in imprecision. Actuaries know that a certain percentage of people will suffer some kind of loss, but the precise individuals who will be affected can’t be predicted. Insurance business models are built around this imprecision; insurance enables a group of people to protect against individual uncertainty by pooling premiums to cover the few of them who wind up suffering a loss. However, as Gupta notes, if those specific individuals could be predicted, there would be no need to socialize risk. Once better predictive models are commonplace, “kaput…goes the insurance industry,” says Gupta, or at least insurance as we know it today.
This is exactly what is happening in marketing. Old models built around imprecision are being toppled by data, digital, algorithms and AI. New models built on better prediction are taking their place. The organizational imperative for future success is to nurture and invest in predictive skills. Skills with data, digital, algorithms and AI are a necessary part of that, but only part of it. Unfortunately, too many companies feel that once they have mastered data, digital, algorithms and AI, they have what they need in place. But they’re just getting started.
Predictive skills also require new metrics, different infrastructure for execution and logistics, a commitment to rapid learning and continuous change, a deeper understanding of probabilistic thinking, comfort with acting and reacting under uncertainty, and a willingness to weed out entrenched cognitive biases built into corporate culture and established decision-making processes.
Most importantly, brands and businesses must ask different questions. Companies should not ask how to make existing models work better in the new world of precision. Instead, companies should ask what new models better fit a world of precision. For example, don’t ask how to make shopping better, ask how to eliminate shopping altogether by knowing precisely what people want. That’s what Amazon Dash Replenishment Service is trying to do. Don’t ask how to better act on what consumers say they want, ask how to better anticipate and predict what consumers want. That’s what the recommendation engines of Spotify and Netflix are trying to do. Don’t ask how to better sell consumers on a product’s features and attributes, ask how to deliver a product or service with a more precise fit to the features and attributes of individual shoppers and consumers. That’s the kind of personalization that many of the new fast-growth start-up brands are bringing to the beauty products category.
Precision will continue to create a crisis as long as marketers are unwilling to move outside the comfort zone of imprecision. The only way forward is to fully embrace the new imperatives of prediction and precision.
About the Author | J. Walker Smith
J. Walker Smith is chief knowledge officer for brand and marketing at Kantar Consulting and co-author of four books, including Rocking the Ages. Follow him on Twitter at @jwalkersmith.