Will A.I. take over marketing?

Hint: not if we’re smart about it.

Admittedly, the cliché that the best, fastest and easiest way to be wrong is to try and predict the future may apply here. That being said, we have been having an increasing level of conversation around the impending impact of A.I. on the future of marketing. More specifically, the conversation covers whether A.I. will change what was historically viewed as a commercial creative outlet for artists, psychologists and philosophers into a heartless statistical algorithm depends on how much you’re willing to embrace technology and how attached you are to what encompasses the business of marketing today. Let’s follow a few emerging trends and technologies into the future and infer how they’re coming together to change the marketing landscape. And finally, let’s explore what marketers can do today to stay ahead and take advantage of the transitions to come.

 

In the beginning: efficiency and margin for all

The first A.I. technologies emerging in the market aren’t job killers, they’re playing field levelers and amplifiers. We are already seeing powerful A.I. tackling jobs and tasks that simply never existed before, because they involve processing and data analysis on a scale that humans simply are not innately capable of handling. Where we used to approach segmentation from an intuitive perspective, now machine learning algorithms are tackling the job with brute force and an unbiased eye. More customized and personalized user experiences are being driven by better product recommendations fueled by massive cluster analysis. Google provides these more finely tailored search results using its RankBrain A.I. Modern fraud prevention has moved from traditional statistical analysis to incorporating natural language processing A.I. to spot suspicious activity. Ad networks rely on A.I. to behaviorally target ads and optimize CPM on the fly. And let’s not forget the current onslaught of A.I. driven, 24-7, low-wage bots handling any number of process driven customer service functions from helping you change an airline reservation or process payments.

As marketers, the advantages come from understanding where the use of current technology can create a more compelling and relevant brand experience for the customer. Traditionally, we created paths for customers, whether physical or virtual, that were designed to help them navigate a linear process of awareness, understanding, affinity and purchase. However, the increased prevalence of A.I. driven experiences is quickly training customers to expect that solutions be delivered proactively based on what we know and predict about their needs. Remaining competitive today will require user experiences that deliver on that expectation. In simpler terms, we used to create great stores. Now, we need to leverage A.I. to create amazing personal shoppers.

 

The next advance: coming for the creatives

If I may paraphrase, the self-comforting statements I hear from many marketers usually go something like, “A.I. will disrupt media, but machines will never be creative.” Even if I subscribed to this prejudice (which I don’t), what machines are great at, especially A.I. driven machines, is learning rules by example and generating output based on those divined rulesets and measured outcomes. That ruleset could easily be as a natural language creative brief or a set of brand guidelines, or, in more advanced cases, self-defined rules based on machine learning applied to a data set of historically effective advertising. In 2016, Microsoft released a short film about the results of an A.I. experiment they titled “The Next Rembrandt”. The film documents how, through a process of feeding the complete collection of Rembrandt portraits into a specialized A.I., they were able to divine the most common aspects of his works, brush technique, lighting, subject matter, etc., then have the A.I. create a “new Rembrandt portrait” from whole cloth, so to speak. The tangible results were astonishing—though arguably neither truly new nor a real Rembrandt. On the more conceptual side, McCann Japan pitted one of its human creative directors against an A.I. that was fed a diet of award-winning television commercials in a contest to create a new TV spot for Clorets Mints. In the end, consumers preferred the human creative directors work, but advertising professionals found the A.I.-generated concept to be more creative. Both anecdotes, while impressive, are still one-offs. The true value will be in the ability for A.I. to generate and iterate thousands of ads for the same product, then optimize deployment around the best performing. No creative pitch. No rationales. Just measurable results.

So, where does that leave us as marketers? The truth is, as with any automation solution, in most cases, A.I. creative doesn’t have to be better than that produced by the best humans, it only has to be good enough to achieve the desired result at a cost substantially lower than a human. And that means we should expect A.I. will soon be capable of assuming what I’ll call production-level execution. Commercial applications like thegrid.io are already applying this idea to web design, to mixed but promising results. The real advantage, in the near term, will go to creative agencies who leverage the associative and generative powers of A.I. as the ultimate brainstorm amplifier. Then apply traditional human curation and execution to the best ideas of the lot—which is exactly what our friends at McCann Japan did, despite their full-on-robot-creative-director PR spin.

 

Infinity and beyond: battle of the bots

Today, you can have a relatively one dimensional, though convincingly polite, A.I. bot named Amy schedule your meetings. Provided by a company called X.ai out of New York, you cc Amy (or, should you choose, Amy’s “male” counterpart, Andrew) on any e-mail message where you are attempting to schedule a meeting, and Amy takes over negotiating the date, time and channel/place. While useful and admittedly quite convincing, Amy is a single purpose A.I. acting within a highly constrained set of rules. And the math is binary. Either requested slots are available or not, and she fills them accordingly. You may have encountered Amy already, and if you ignored the small disclaimer at the bottom of her emails, you could easily be fooled into becoming jealous of your business acquaintance’s ability to hire a personal assistant. But if you had pressed the conversation, Amy would have quickly succumbed to the Turing-test realities of her limited purpose. The more interesting aspect of this technology, however is what happens when two people who want to schedule a meeting both use Amy. At that point, the emails stop and the two bots “go binary” and negotiate, invisibly, in the background and only alert if a conflict is seemingly unresolvable. That is a more apt model for imagining how A.I. will manage our consumer interactions. For marketers, preparing for competing in this inevitable future means developing data rich user profiles that allow for predictive behavioral modeling and precise segmentation, as well as creating fully-abstracted API-level transactions that can easily manifest in voice, bot, or human driven experiences.

 

It’s not all gloom and doom!

If history has taught us anything, it’s that technological progress favors those most adaptable to change. Someday, marketing might be a hands-off, A.I. driven cloud service, as emotionally indifferent as an email server, but not in the near term. As marketers, we shouldn’t focus on being replaced by A.I., but rather on being overtaken by human competitors who leverage the technology to connect more meaningfully with our customers.


If you’re looking for intelligent ways to enhance your marketing strategy, integrate new technologies, or improve your customer experience, contact us. Or, click around our site to see all of the other ways we help we help our clients outsmart, outmessage and outmarket larger competitors, to steal their most profitable customers.