AI Still Can't Create the Perfect Marketing Campaign for You Gathering data and processing purchase signals is one thing, understanding context is another.
By Sanjay Castelino Edited by Dan Bova
Opinions expressed by Entrepreneur contributors are their own.
Author Dave Eggers scared millions with The Circle, his tale of AI-driven surveillance that's since become a movie starring Tom Hanks and Emma Watson. And why not? Decades of science fiction have trained us to believe machines will eventually replace humans in everything -- even when it comes to marketing.
Related: 10 Ways to Use Chatbots for Marketing and Sales
There is some basis for this belief. Machines have already replaced some learnings to make marketing more efficient. For example, online algorithms like the ones used in Google ads have made it easier to determine purchase intent and predict when prospects are ready to buy. Machines are brilliant at processing data that represents purchase signals. But, what they miss is context about buyers' existing brand perceptions and the motivations behind their purchase decisions.
Context: The key to creating meaningful brand experiences
Anything that informs motivations is context. Let me give you a personal example. I may be searching for signs of a failing hard drive or for good deals on a lightweight PC. To Google, these searches will likely signal that I'm in the market for a computer. What sort of computer will it choose to show in an ad?
Maybe Google knows I'm in marketing, I work at a tech company and I live in a creative hub -- Austin, Texas -- all of which might increase the odds of me being in the market for a Mac. What kind of Mac, though? A check of my search history likely reveals I frequently hunt for good airfares when traveling for business, which leads Google to present me with travel-friendly laptop options.
A new, souped-up MacBook sounds great ... for just about anyone else but me. I've just never been an Apple guy; my brand loyalties lie elsewhere, which is crucial context when presenting me with an offer. Despite the purchase signals I was sending, my brand preferences were overlooked and the ad was essentially a waste of time. So, how can we expect AI technology to create and execute a full marketing campaign if today's algorithms aren't even equipped to display an effective ad?
Related: AI Is Taking the Art Out of Sales
The challenge for marketers: Injecting context, interpreting data
That doesn't mean machines that process data and interpret buying signals are unimportant. But, they still lack context about buyers' existing brand perceptions and whether or not those perceptions can be influenced. They also lack the ability to define what experiences should be created to reach buyers that are receptive to your message.
Sure, machines can provide data to help marketers find prospects that might be in the market for their product. They can also help inform micro-decisions to more efficiently run and promote campaigns. But, they're not equipped to create personalized brand experiences that can grab your attention.
To effectively reach and engage buyers, an understanding of their current perceptions and past experiences with your brand is required. That's why the creative exercise of developing impactful experiences that take prospects further down the funnel, from awareness to consideration, is still better informed by marketers than machines.
Related: Chatbots Can Help Your Team But They Can't Replace Your Team
Artificial intelligence for a data-driven future
But, if you think about it, context is just another layer of data so there could be a future where machines expand their algorithms to better understand how customers and prospects perceive brands and consume experiences.
For example, let's say you're launching a new brand campaign to reaffirm your commitment to customers. Using a model that examines certain behaviors, such as browsing history, Google searches, social media interests and more, AI technology might be able to tell you which customers are more likely to jump ship to a competitor and need a little more attention. But, what experiences should be created to effectively reach the customers at risk? That question should still be answered by a human being.
Plus, the privacy implications for harnessing this sort of data for AI are vast. Look at Facebook for example. What if my likes and dislikes, and the time and place I clicked on content that interests me, were made public? What if my friends that share my interests were made public as well? Suddenly, machines would be infused with an extra layer of context -- context provided by the personal elements of my life.
Related: Survey Finds Marketers Know All the Important Tech Trends But Aren't Prepared for Any of Them
But, I might not like that, and it may encourage me to share a manufactured or curated version of myself while hiding the parts I want to remain secret. Or, I may choose to stop sharing anything at all.
All of which brings us back to the human experience. Machines are capable of ingesting and processing and making some limited meaning out of more data today than ever before. That's good, but it may not get us as any closer to our customers.
Why? Because data and signals are nothing without context, and context is informed by our existing perceptions and motivations that can be easily hidden. Until AI can perfectly replicate the artful act of human engagement and then make educated decisions about tapping into those motivations, marketers will not be replaced. Today's machines can't do all that, and it'll likely to be years, or even decades, before they can.
Related Video: The Only Online Marketing Metric That Actually Matters
