In July, JPMorgan Chase made an announcement that struck fear into the heart of many a creative: They’ve signed a five-year deal with Persado, a company “reinventing marketing creative by applying mathematical certainty to words.”
Do what, now?
It does my job, but with artificial intelligence.
How this all started
Chase started with a three-year pilot program testing the effectiveness of Persado’s AI writing digital copy for their card and mortgage services. The resulting creative has garnered a higher percentage of clicks than comparable human creative — sometimes twice as much. Compare a human-written digital ad:
Access cash from the equity in your home.
to the AI’s version.
It’s true — you can unlock cash from the equity in your home.
The scary thing is? The computer’s version is actually kind of better.
Chase has been so satisfied with the results that it’ll be applying Persado’s AI across all its service lines. And Persado says it’s already working with 250 brands across the country, including some pretty big names, to expand the use of AI in marketing creative.
How machine learning works
So how would an AI copywriter work, anyway?
A vastly oversimplified and probably not entirely accurate overview, because I am not by any means a software engineer: Artificial intelligence works via a computer algorithm that’s fed by data and constantly refined by its own successes and failures. This could be through supervised learning — No, algorithm, bad or Good job, algorithm! That’s exactly right — (which doesn’t necessarily need human supervision) or through unsupervised learning, in which the algorithm finds hidden patterns and structures on its own. It’s trial and error at four billion cycles a second, and it doesn’t need your help, thank you very much, it’s got this.
(When Google asks you to prove you’re not a robot by selecting all the images that have a storefront in them, you’re actually helping to “train” an algorithm by showing it what a storefront looks like and doesn’t look like, so it can better identify storefronts in the future. That’s right — by proving you aren’t a robot, you’re actually helping to train a robot to be better at pretending to be human. This is the world we live in.)
Essentially, it’s the Infinite Monkey Theorem on NZT-48. And in this case, the typewriter-wielding monkeys are after my job.
Some more entertaining examples of machine learning include the fun antics AI researcher Janelle Shane has been having with her neural network, teaching it to do things like create recipes and name colors. (Highlights include Sane Green, Snowbonk, and a lovely Kelly green named Bylfgoam Glosd.) There’s also this example of Google’s DeepMind AI learning to walk, which I could watch on loop for hours.
The thing is, though, as silly as that video looks, it represents something huge: continuous improvement in an automated system that, given enough time and the right feedback, will eventually learn to be Usain Bolt. And Persado has been at this for seven years.
Why I’m not scared
Well, okay, I’ll admit to being a little bit disappointed. This whole time automation has been posing a growing threat to meatspace employment, I’ve always cavalierly laughed, “Ha ha, at least a robot can never do my job!” So naturally, a marketing firm is going to come along and say, “Hold my beer.”
Realistically, the real threat of automated copywriting isn’t that the AI-powered creative might be superior but that it might be good enough for potential clients — that the creative is good enough, convenient enough, or just novel enough for a brand to go, “Ooooh, robot,” and suddenly, I’m out of a job. And while that might sound like reason for concern, it’s actually kind of comforting. Losing out on a gig to an AI is no commentary on my skill as a copywriter. I can’t out-creativity a competitor who’s being assessed on metrics other than creativity.
And Persado’s CEO and co-founder, Alex Vratskides, does identify the value of the company’s AI not as a producer of superior copy but as a producer of more accountable copy. “To the creative community, the marketing community, this brings accountability and data-driven insight,” he told AdAge. “If you go to any marketing creative out there and you ask, ‘How did you come up with that, why did you use that word and not that word,’ they cannot actually answer. With Persado, there is a mathematical answer.”
While advertising creatives will never be able to provide the mathematical accuracy of an algorithm, we can always come to be more conscious of the reasoning for our creative choices, and we can be better about articulating that to clients. We can stay abreast not only of the creative skills and trends that keep our work fresh but also the psychology and sociology of consumer sentiment to inform our work (which we should be doing anyway, y’all).
Chase has said they’ll use Persado at the ideation stage of copywriting, working alongside — not as a replacement for — the bank’s marketing team and outside agencies. While an AI might be able to pick out the perfect word to drive click-through on a digital ad, it takes more than good words to make a good concept, a good campaign, a good event, and an algorithm will never be able to match a good creative team on that front.
Well, not never. These things are literally, by definition, continuously getting better. But it’s not going to happen in my lifetime, and after that, who cares? I’m dead. Best of luck, baby copywriters.