Blog post

Personalized Marketing in the Age of Weaponized Data

By Andrew Frank | April 02, 2018 | 1 Comment

MarketingMachine LearningDisruptionData-Driven MarketingAIAdvertisingDigital Marketing Strategy and ExecutionMarketing Data and AnalyticsMarketing Technology and Emerging Trends

The pace of events unfolding in the volatile world of data-driven marketing and privacy has outstripped the market’s ability to process them. Last week’s Adobe Summit in Las Vegas, a monumental event many months in the making, exposed some of the tensions facing marketers and technology providers as they struggle to come to terms with shifting attitudes toward the use of data for marketing. Adobe had two main messages for its audience of roughly 12,000 attendees: become “experience makers” (“people buy experiences, not products,” said CEO Shantanu Narayen in his keynote), and embrace the “holy grail of hyper-personalization at scale.”

These messages were clearly meant to be complementary and consistent, but the second one struck a faintly dissonant chord with news unfolding at the time. The next day, Facebook announced plans to shut down its third-party data provider division as part of a wider effort to address fallout from its Cambridge Analytica scandal. The same day, Adobe announced general availability of its Experience Cloud Device Co-op, a service it developed to allow marketers to share information relating devices to individuals for targeting purposes – with industry-standard privacy protections in-place. Meanwhile Acxiom’s CEO, responding to Facebook, wrote a blog post titled, “A World Without Walls” to make (again) the case for ethical use of data without the restrictions of walled gardens. As all this was happening, curious Facebook users all over began downloading their profiles after Facebook released a new privacy page. (Here’s what some of them found.)

All in all, it was an eventful week.

Kalev Leetaru, writing for Forbes, put it like this: “All in all, it is remarkable how in the space of just 8 years data has gone from hero to villain in the world of campaigning….” Although he was writing about Facebook and politics, the sentiment may extend to all data-driven marketing. Nowhere is the contrast more apparent than in Cambridge Analytica’s own shifting messages, from extolling the amazing power of its “secret sauce” to insisting that, “no data from GSR [the contractor that obtained Facebook data] was used by Cambridge Analytica as part of the services it provided to the Donald Trump 2016 presidential campaign.” Once treasured by enlightened marketers everywhere, data – specifically, personal data – has become trade’s plutonium: powerful and dangerous.

The questions soul-searching marketing wonks (whom I can attest are generally not monsters) are asking is, how powerful and how dangerous? And, how can we fix this?

Back in 2012, I gave a presentation at Gartner’s annual Symposium titled “The Rise of Influence Engineering” (subscription required). It explored a hypothetical future where psychology bridged the gap between data science and marketing, empowering machines with tools of “built-in automatic response” as Cialdini called it. Has it happened? The consensus warrants great skepticism regarding the effectiveness of the psychographic techniques in the news these days. But there’s no denying the writing on the wall: marketing technologists, armed with machine learning and reams of personal data, are now in active pursuit of techniques of persuasion that attempt to exploit the observations of social psychologists and behavioral economists like Cialdini, Thaler, Kahneman and Tversky who have collectively revealed the vulnerabilities of human decision-making.

Yet, there are a few reasons to be optimistic about all this. As modern advertising’s founding fathers, David Ogilvy and Bill Bernbach, were both known to remark, effective advertising will accelerate the death of bad products, and social media amplifies this effect. People learn, and if they are repeatedly tricked en masse into making decisions they later regret, they will adapt (many will anyway). Perhaps that’s what we’re witnessing: a slow dawning social awareness of the unforeseen consequences of sharing excessive personal data on social networks. Privacy flare-ups, like gun violence, have been the periodic norm for years. This one somehow feels different, although changes will take years to play out. As I see it, the changes won’t be so much about impeding the overall flow of data as they will about people wising up to how to recon with its more manipulative applications, both in marketing and politics, and exercising more control and sensitivity. On the marketing side, we’ll need to learn to do more with less.

The complex relationship between machine learning, privacy, and marketing is among the key topics on the docket for next month’s Digital Marketing Conference. I’m excited to be presenting on topics such as What AI Teaches Us About Being Human, and How to Cooperate and Compete With Google, Amazon, Facebook and Apple. Sorry for the plug, but hope to see you there.

Comments are closed

1 Comment

  • Mike says:

    very thoughtful post, Andrew. I think this is the year that “hyper-personalization at scale” starts to sour in the mouths of many marketers when confronted with the many pitfalls associated with trying to achieve it AND balance it against regulatory forces and consumers’ general unease with what amounts to the “marketing surveillance state.”