Buzzwords. So many buzzwords. I mean, I get it – I’ve worked in marketing my whole life. And while buzzwords can be useful signals, shorthand, or common language, they can also be dangerous. And we are in a dangerous time for marketers. It’s not quite the days of Skynet, but more and more marketers are being duped by a particularly egregious buzzword: artificial intelligence.
Artificial intelligence (AI) is, to put it bluntly, not what you think it is. It is not (and never can be) a tool, or a product, or a feature. AI is only an outcome, and any platform or product that suggests otherwise is likely lying to your face. So while you’re enamored with the idea of an intelligent, sentient machine designing, deploying, and optimizing your marketing program, the results won’t be satisfactory, because it literally does not exist.
Let’s talk briefly about what artificial intelligence actually is, and then talk about how it’s marketed and why that disconnect is failing marketers. And then let’s talk about how we can get better together.
We can use machine learning as a tool to achieve artificial intelligence, but it’s not a prerequisite for those outcomes.
First and foremost, artificial intelligence describes an outcome. More specifically, AI is better described as being achieved when a machine performs a task that simulates, mimics, replicates, or displays characteristics of human intelligence. A complex example of realized AI is image classification, in which a machine correctly identifies the object in a photo. A much simpler version would be a macro that transforms raw data into a dashboard without user intervention. Both are tasks humans can easily do, and we’ve either trained an algorithm or coded a macro to have a machine do it for us.
But AI is not a tool, nor is it a product. In fact, it cannot be either of those things. Instead, we can use machine learning as a tool to achieve artificial intelligence, but it’s not a prerequisite for those outcomes. Machine learning itself is another discussion. It requires a lot of legwork for it to be successfully implemented. From defining the problem to exploring the data to deploying a meaningful algorithm. The machine just doesn’t learn in a vacuum.
So what does this mean for you as a marketer?
1. Be wary of any tool that claims to be driven by artificial intelligence
If they are using this marketing buzzword to describe their technology, it stands to reason that they may not really know what they are talking about.
2. Question any solution that appears to be magic
Mimicking human intelligence is incredibly hard. Any turnkey solution is unlikely to provide you the results you’re being sold or you might anticipate, since it inherently has to be broader than the business objective you’re likely trying to solve.
3. Work backwards from your objective
Machine learning solutions to achieve artificial intelligence can be extremely powerful, but they can also be woefully misused. Your business objective may be solved using another technique, method, idea, or solution, and those solutions can only be dictated by a clear understanding of what you’re looking to achieve. Don’t jump to the solution before you understand the problem.
4. Arm yourself with knowledge
Companies and products touting AI-powered solutions can be an extremely shiny object. It is appealing to discover an optimized process that can automatically – through technology – solve really complex problems. So it’s more important than ever for marketers to arm themselves with enough working knowledge of these emergent technologies and what they actually entail before making a poor decision and wasting time, energy, and most importantly, budget on a solution that is nothing but buzz.
Jacob Davis is the Senior Director of Strategic Services at Cheetah Digital. He is a cross-channel marketing professional with over 5 years working on evidenced-based strategies and analytics. Jacob leads a team dedicated to providing clients insight into their customers & campaigns using historical data, advanced analytical techniques, statistical methods and predictive modeling. Based in New York City, he works closely with enterprise brands, across all industry verticals, to help derive unique, nuanced insights and data products that allow for better marketing strategies and decisions. Prior to working at Cheetah Digital, Jacob spent time in developing product for Abercrombie & Fitch, doing land analysis for a subsidiary of Shell and consulting with a variety of small businesses.