A I On Every Marketing Team
With driverless trucks being tested all across the country, and IBM’s Watson making vast life saving improvements to doctors’ diagnoses, it is time to be honest about how vulnerable human marketers are to the looming AI explosion. This is not a doomsday scenario. Nuclear warheads do not obliterate Earth’s outer crust at the conclusion of this article. However, there will be some aspiring marketing professionals who feel like their 10 year plan is about to collide with a WMD.
The first harbinger of any good technological apocalypse appears as a friend. A Trojan Horse that turns your 60 hour work week slog into a 25 hour work-from-home romp. You’re more efficient, you’re more effective, and status reports practically write themselves… because in an AI supported work environment, they do.
Imagine how quickly you’d adopt a bot that could scan engagement on your posts. If you’re like me, dozens of hours a week could be saved if you had someone else find the replies that required attention. No more hours burned scanning endless innocuous messages. As long as the AI had a high enough accuracy rating, I would happily pay for access to that tool.
Imagine starting a blog post. Instead of staring at a blank Google document for an hour searching for your muse, you’d turn to a blogging research assistant. Simply tell the AI what kind of post you want to write. Five seconds later, the computer provides research on the subject, with links to where the info was pulled from, so you can verify the context. After using this autonomic assistant diligently for a few months, it could start suggesting subject matter based on your previous interests and recent articles. Sound to sci-fi? Netflix does this with its rating system already.
How hard would it be for an AI to start making suggestions about changes in word choice? A quick thesaurus search filtered by the reading level of your audience, then improved by analysis of engagement on previous posts, is just one of a dozen possible recipes us marketers might use to optimize our copy. Maybe you ask for a comparison of nut graphs from articles similar to the one you’re looking to write. Maybe you want keyword suggestions that your target audience has a high probability of searching for, that relate to your post.
Now apply this imagined technology to the task of writing ad copy. The task becomes a breeze if you can quickly find similar ads to draw from and test them against high value demographics. But that kind of research would take a human being days to put together. A near-future AI trawling through Facebook ads could return color or iconography suggestions in real time. Now that driverless vehicles use computer vision to navigate, it’s only a matter of time when advancements in the field start leaking out to other industries.
The above examples are not speculative. Research and testing is the bulk of what marketers do. Sure, our rough drafts come from the heart, but the final results are always filtered by practicality. Did the ad work? With who? How well?
In the end, what we create is far more influenced by the facts than most of us admit. Facts about our target audience. Facts about the current climate of the market. Even facts about our product. Today, AI can leverage facts well enough to drive a car through the physical world filled with chaotic human drivers. Today, AI can diagnose and recommend cancer treatments on par with elite medical centers. Marketers are not far as far behind the wave of disruption as we like to think. The creative side is part of what we do, but it’s not the majority.
The marketing A.I. apocalypse doesn’t happen in a climactic explosion or the sudden collapse of the industry. It starts when the workload gets light, it takes less people to do the job and there’s a serious lack of busy work for the interns to do. Sounds nice.