Months ago, a colleague forwarded me a link to a company that claims to have found the answer to every marketing team’s content creation woes: robots! The company says that, with the help of artificial intelligence, it can “generate” and “optimize” your marketing language (but not rid it of jargon, apparently).
I was skeptical.
Can the AI bots interview a reticent CEO, pull the best bits from the conversation, and wrap those few gems into months’ worth of thought leadership content? Can the bots craft articles and blog posts that do more than hit the right keywords for other bots to find and deliver through search engine results pages? Can AI do the much harder content work of engaging readers in a way that instructs, persuades and generates leads among real-life humans with the power to make budgetary decisions?
I was hoping the answer to all those questions was either, “Not really,” or better yet, “Hell, no!”
Obviously, some of that hope stems from an instinct for self-preservation. But content marketers (along with many other professionals) aren’t being paranoid when they fear AI competition for their jobs.
In a recent op-ed for the Washington Post, Bryan Dean Wright notes why workers across professional spectrums have reason to worry. He writes, “…corporations and investors are spending billions — at least $8.5 billion last year on AI, and $1.8 billion on robots — toward making all those jobs replaceable…Powerhouse consultancies like McKinsey & Co. forecast that 45% of today's workplace activities could be done by robots, AI or some other already demonstrated technology. Some professors argue that we could see 50% unemployment in 30 years.”
This topic came up at a LeadBoston program day I attended recently, where public school leaders noted the AI future is something they consider while designing curriculum for today’s students. If almost half of what workers do today – from driving busses to giving financial advice to writing marketing content – will eventually be automated, how do you prepare the next generation for working adulthood? Will working adulthood look anything like it does now?
Human content for human buyers
These are heavy, existential questions, but I feel a lot better about the threat for potential obsolescence when I see what AI is producing in the realm of content marketing today. When my colleague pointed out the AI content creation tool I mentioned above, I signed up for the company’s email newsletter. “Let’s see what they can do,” I thought.
In every month since, a message has landed in my inbox that makes me confident about the continuing necessity of human writers in the world of marketing. From this unscientific sampling, I’ve gathered that AI “writers” have been programmed with some sort of emoji-laden dictionary. Their language is heavy on gifs and bro-speak, and light on substance.
All of that would be fine…if the target market for content were also robotic. It’s not. If you’re trying to connect with human decision-makers, nurture potential human customers and persuade them that your point of view is valid, it takes more than 100 iterations of a gif-laden template to get the job done.
Not only does effective thought leadership still require thought (i.e., the thoughts of a real person with an actual human brain and human experiences), it requires craft. The people sitting down at their inboxes every day deciding whether to click “open” or “delete” are tired of robotic content. They are tired of how-to articles that don’t really teach them how to do anything. They are tired of top 10 lists that are clearly scraped from some other website (which scraped them from other websites, which scraped them from other websites, ad infinitum).
Content marketing shouldn’t make your prospects tired – or skeptical, annoyed or angry. Content done well should fire people up, because you taught them something they needed to know, told them a good story, made their jobs a little easier or pushed them to consider a new idea.
There is a (limited) place for AI in content marketing
Some kinds of content are so structured and pre-defined that they’re easy to delegate to an AI workforce. An earnings report, for example, is not a great place for a writer to exercise creativity or the art of persuasion, and there are natural language generation tools on the market that can successfully tackle these formulaic content types.
Journalists have been busy testing out this premise. At the Financial Times, writer Sarah O’Connor went head-to-head with “autonomous artificial intelligence” and found that AI produced factual content faster, but failed “to distinguish the newsworthy from the dull.” And at the Washington Post, an AI tool called Heliograf is “writing” stories based on structured data. The director of strategic initiatives for the newspaper, Jeremy Gilbert, recently told Wired that AI frees the editorial staff up to focus on more challenging work.
Wired quotes Gillbert as saying, “If we took someone like Dan Balz, who’s been covering politics for the Post for more than 30 years, and had him write a story that a template could write, that’s a crime.”
In marketing, as in journalism, there are areas where technology can take repeatable tasks off human to-do lists, thus freeing workers to focus on areas where our abilities still beat the bots. In analytics, segmentation, personalization, multivariate testing, distribution and more, automated tools allow marketers to increase efficiency and efficacy. And, yes, when it comes to writing formulaic content based on structured data, marketers can leverage AI.
There is a place for this template-like content in marketing, but it should be a limited one. At the core of any content strategy worth executing, there must be helpful, human content worth the time it takes recipients to read. In a landscape so littered with mediocre content, why would we want to create more, no matter how efficiently we could churn it out?
When it comes to meaningful content creation, hiring humans remains a powerful differentiator.
Ready for a (human) partner who can push your business forward?
