Everywhere you look, artificial intelligence is being touted as a solution for just about all digital aspects of businesses. It seems like AI is even the answer to some questions we aren't asking (AKA, solving problems we don't have).
Marketing is not immune to this push for AI. Quite the opposite, actually. Trade publications are inundated with references to, and hype for, artificial intelligence. And there's no denying all the data derived from digital marketing opens the doors to useful applications. The term "data-driven" has been circulating around the industry for the past decade.
Of course, marketing has always been data driven to some extent, but our abilities to uncover insights have been enhanced exponentially by the digital transformation. Artificial intelligence is the progression, the next step, of this data science proliferation for marketers, but there seem to be some important caveats lurking beneath the hype.
Vendors of AI solutions often tout remarkable results. Though some marketers may have indeed benefited from implementation of "AI" applications, many others have seen tepid results at best.
Some of those shortcomings could be attributed to poor or insufficient data, but some of it is because we're swimming in some precarious waters of marketing technology. We're being oversold on the impacts of AI marketing technology, and that could leave buyers with a snake oil view of it; that, in turn, threatens to engender a skepticism that might be difficult to overcome when the currently impossible in fact does become possible.
True artificial intelligence, at its foundation, is teaching a computer to mimic human behavior. Current reality is that computers are great at doing many tasks the human brain does poorly, or slowly, such as aggregating and segregating; they can draw assumptions based on massively scaled datasets.
Yet, though computers may be excellent at replicating intelligent behavior, they still come up short when asked to think intelligently.
Not There Yet
Consider that mankind still cannot fully understand how the human brain develops thoughts, so how can we train machines to develop intelligent thoughts? At the end of the day, AI is still subject to the behavior humans design into it, and that can lead to unwanted outcomes. Just ask Amazon.
In 2014, engineers for the retail giant sought to automate the company's hiring process by creating an AI that would find ideal job candidates from around the world. Amazon soon discovered the system was biased against women. The flaw was in the algorithm, and even though developers tried to iron out the kinks, they failed. The project was abruptly discontinued in 2017.
Startups and industry giants alike have attempted to put advertising into the virtual hands of machines. Saatchi & Saatchi employed the famous IBM Watson to write copy for some advertising. Chase Bank is using an AI platform to tweak some of its marketing copy.
On the other hand, many in the industry maintain their skepticism, as highlighted in Burger King's ruse on AI-written advertising. Though it's parody at AI's expense, the Burger King satire highlights a disconnect within the marketing industry about AI's potential.
Business-to-Business Still Means Human-to-Human
Artificial intelligence and machine-learning hold promise for marketers and advertisers, no doubt. We've probably seen only flashes of what might be possible in years to come. But, still, and probably for some time to come, the best marketing—especially business marketing—relies on old-fashioned human intelligence.
Developing brands and the relationships that are so important in B2B businesses require human and emotional connections. Powerful computers can crunch massive data pools and determine the best times to serve ads. They can be trained to make smart media purchases through real-time bidding. But they can't understand how or why emotional connections happen, and those emotional connections have always, and still remain, a vital part of business-to-business.
The future of martech need not be completely artificially intelligent. Automated processes can be helpful, but believing that AI implementation can or should replace human intelligence (and, more important, emotional intelligence) is doing the industry a disservice.
B2B has been transformed beyond its handshake-agreement days, but not beyond the need to maintain the human intelligence that forms those relationships. People make those connections happen, and therefore people will always play a major role in the development of your business's marketing strategy.