At its very best, advertising is art. Everyone knows that. During the industry's golden age, marketers and agencies routinely employed illustrators as diverse as Ben Shahn, Andy Warhol and Norman Rockwell to furnish their campaigns with creative of the highest aesthetic caliber.
Less understood is the longstanding relationship between scienceâ€”both social and physicalâ€”and advertising. It was trained cognitive psychologists such as George Gallup, a professor at Drake University who later started Young & Rubicam's first research department, and Paul Cherington, director of research at J. Walter Thompson Co. and professor of marketing at the Harvard Business School, who in the 1930s first developed methods for segmenting and targeting large and previously undifferentiated consumer populations.
For roughly 80 years starting in the early 1930s, agencies grew increasingly sophisticated about targeting audience segments across largely static mediaâ€”broadcast (and later cable) television, radio, out-of-home, print and even early digital (i.e., the World Wide Web). Indeed, segmentation was a core competency in an era of static media, and social science added tremendous value to the life of a campaign.
But we don't live in an age of static media anymore. We live in the age of the programmable internetâ€”a dynamic, highly personalized network of interconnected devices, applications and consumers. The programmable internet delivers a bespoke experience to a billion consumers. Just compare your Amazon or Pandora apps to someone else's, and you'll see what I mean.
Today, advertisers need data science to harness the power of their art. That means building algorithms that enable the delivery of bespoke advertising experiences. Simply put, if advertisers can't deliver campaigns that are every bit as customized as digital content itself, they can't drive brand or performance at scale.
If you're an e-commerce site specializing in wines and spirits, you're likely able to program your own page or app to deliver customized marketplaces to millions of individual customers. You've forged relationships with these customers. You can anticipate their needs and interests and propose new ideas and opportunities that you think they'll appreciate.
If you're a brick-and-mortar clothing retailer that also has a digital marketplace, you're interacting with your stakeholders in multiple locations and contexts. You're in possession of powerful point-of-sale data that you already use on your owned properties. But you need to extend this bespoke experience beyond your owned properties. You need to deliver a bespoke marketing experience.
If you're a b-to-b service provider that relies on third-party data services to arrive at a deeper and more meaningful understanding of your stakeholders and potential stakeholders, you need a feedback loop that enables you to market and learn simultaneouslyâ€”because the programmable internet evolves in real time. It's not staticâ€”it's dynamic. And so, too, should your marketing program be dynamic.
The good news: The world has lots of data scientists who know how to build great algorithms.
The bad news: Most algorithms are neither proprietary nor portableâ€”meaning marketers are usually locked into using black-box algorithms. They can put their data in, but they can't pull their data out, and when they port their business to a new platform, they're locked into yet another black box.
Marketers need a programmable platformâ€”first, to create and plug in their own algorithms, and, second, to port those algorithms to other platforms should they decide to make a change.
This is the challenge for marketing and ad tech. If there is no longer such a thing as a one-size-fits-all internet, then a one-size-fits-all algorithm no longer suffices when it comes to consumer engagement.
If you're Ford, and you're trying to compete with General Motors, for example, and you both use the same black box algorithms, your only points of differentiation are your data and creative. By building and owning your own algorithms, you have a new, powerful way to forge bespoke conversations with potential consumers and thereby beat the competition.
Much of today's marketing infrastructure doesn't support proprietary and portable algorithms. It needs to.
The call to action for every agency and marketer is this: Demand that your technology platforms have the ability to control your destiny through data. It takes an artist to make compelling creative. It takes an algorithmâ€”scratch that; it takes your algorithm to customize creativity at scale.
About the Author
As co-founder, CEO and chairman of the board of directors, Brian O'Kelley leads AppNexus' strategic initiatives. He has more than a decade of experience in online advertising, including CTO of Right Media (later sold to Yahoo!), where he invented the world's first advertising exchange. Brian is a contributor to Forbes on technology-related topics, has been named to Crain's 40 Under 40 and Adweek 50 lists, and was recognized as an Ernst & Young Entrepreneur of the Year in the New York region.
Brian holds a B.S.E. in computer science from Princeton University. He lives in New York with his wife and daughter.
About the Sponsor
AppNexus is a technology company that provides trading solutions and powers marketplaces for Internet advertising. Its open, unified and powerful programmatic platform empowers customers to more effectively buy and sell media, allowing them to innovate, differentiate and transform their businesses. As the world's leading independent ad tech company, AppNexus is led by the pioneers of the web's original ad exchanges. Headquartered in New York with 23 global offices, AppNexus employs more than 1,000 of the brightest minds in advertising and technology who believe that advertising powers the Internet. For more information, follow us at @AppNexus or visit us at www.appnexus.com/en