How to Tell When a Product Is Truly Powered By AI

With technology moving as fast as it is, we’re all scrambling to stay relevant. And marketers are no exception. Every day, we’re met with stories about robots taking over our jobs and AI-less businesses soon being left in the dust. And in this dramatically shapeshifting panorama, it’s easy to get all shifty-eyed and assume your competitors are nailing the whole AI thing, poised to bury you. So, what do you do? Naturally, you look for solutions to put your marketing capabilities on steroids, vowing to tap the power of AI. And that’s where it gets confusing. Just as marketers are sweating to stay on the cutting edge, so too are marketing technology companies—who then task their own marketers to package up some pretty and aspirational buzzwords. As the knee-jerk reaction kicks in, many of them slap “artificial intelligence” labels on products that don’t truly have the capabilities one might think. “Vendors have done a good job of muddying the waters by providing vague AI solutions with grandiose names that don’t seem to deliver many use cases,” warned Austin Miller, director of product marketing, Oracle Marketing Cloud. “Even the use cases that are supported tend to be statistical models rebranded as artificial intelligence.” You see, it’s easy for marketers to fall victim to clever marketing too. So, if you’re on the hunt for a marketing engine fueled by artificial intelligence, you need to know how to spot what is really AI and what is simply jargon. How can you tell? As Jason Carmel, global chief data officer for Possible very simply put it: “Any time marketing technology makes a decision that a human normally would, it’s being driven by some form of machine learning.” Notice how he used the word “decision” when referring to the technology. The machine is making the call in real-time. Not you. That’s an important differentiation. “This is different than merely setting up rules, which is manually driven by a human and executed without evolution by a machine," Carmel added. "If a marketing technology ingests data and then decides to change something pre-existing, from content displayed to media purchased, then there is an artificial intelligence at play.” Personalized messages can be delivered to a larger volume of audience segments with AI-powered mar-tech. There is one caveat, though, advised Stefana Rusu, senior director of marketing at Acxiom. “These audience segments all require 'good data' to ensure that AI does not fall short and deliver poor results (i.e., garbage in, garbage out). Before feeding data into an algorithm, organizations need to have processes in place to confirm that consumer data is accurate, up to date, and in compliance with standards for ethical data use.” Ask the right questions If you’re in the market for AI-driven marketing technologies, be smart and don’t fall victim to, well, marketing. The most brilliant marketers hawking infantile AI products are very good at deflection. “A good way to zero in on the reality of the situation is to admit to being ignorant on the subject and ask a bajillion questions,” said Carmel. “When a vendor refers to 'algorithms,' making 'smart decisions,' for your business, ask them what those algorithms are and which specific decisions are made, and what the readout is for a human to validate that. Ask them for examples of business using that technology currently and, even better, ask to talk with them.” Carmel confirmed: “I would implore marketers to focus their research on how exactly the AI works. Ask how the answers or results are being generated.” Beware of phrases like, “Trust me, it’s AI.” Seriously. Don’t be afraid to sound stupid When you’re a professional marketer, you might fall into the trap of not wanting to look like an idiot and therefore not speaking up. You might assume that because you don’t understand it then it’s because you’re not very savvy. And you should know everything. You’re a marketer after all, right? Wrong. If something sounds weird, there’s a chance it might be. Make sure you truly understand the tech and that it makes sense. “Ask what data the AI is working off of,” suggested Miller. “This piece is often overlooked, but critical—data is the fuel that the algorithm engine runs off on. If you have bad or absent data, there is a good chance your AI is going to be generating bad results. As with most things in marketing, it all comes back to the data.” If you’ve been tasked with streamlining your marketing department for the age of the robots, breathe deeply before taking the plunge. Don’t let your ego get in the way and remember that there are no stupid questions, especially in such uncharted terrain. At the end of the day, you’ll be the one having to demonstrate what your budget went on and how it works. If you can’t explain what your AI software does, you shouldn’t be buying it for your company.