Marketing’s Artificial Intelligence Revolution Is HereBy Dillon Baker March 14th, 2017
Chief marketing officers have one of the hardest jobs in the corporate world. According to research by the consulting firm Russell Reynolds, it’s only getting harder. Last year, CMO turnover reached its highest point since Russell Reynolds began tracking the data in 2012.
After years of swift technological development disrupted marketing to its core, CMOs and their teams are now responsible for a bevy of critical company functions: optimizing the sales process, improving customer experience, and marketing the product across a complex array of channels. Meanwhile, CEOs understand that technology allows for accurate tracking of marketing dollars spent, creating immense pressure on CMOs to perform.
Now, another technological breakthrough is on the horizon: artificial intelligence (AI). According to many experts, AI has the potential to be the most revolutionary technology since computing itself—and marketing is first in line for disruption.
Should marketing leaders panic or celebrate? It depends on who you ask.
The artificial intelligence hype machine
2016 was a big year for artificial intelligence.
Google’s AlphaGo program defeated the world champion in the complex game of Go, a landmark in AI development. Facebook, Microsoft, Google, and Amazon all made significant investments in AI. Wired proclaimed that the world’s biggest tech companies are “remaking themselves around the technology.” In marketing, Salesforce and Adobe announced new intelligence programs, called Einstein and Sensai, respectively. We’re in the midst of an AI arms race.
“I’m astounded in how AI is going to transform every industry,” said Leslie Fine, VP of product at martech giant Salesforce. “Companies today face an imperative to integrate AI into their products and services, or risk becoming less competitive with companies who are applying AI to improve customer experiences and make more intelligent business decisions.”
But what is AI exactly? And how does it apply to marketing? The answers aren’t as simple as you might think.
Everyone seems to have a different definition for AI. Typically, though, it’s an umbrella term that refers to software that carries out tasks that would normally require human intuition. Think recommendations on Netflix, Google Translate, software powering self-driving cars, and even GPS map services such as Waze.
Marketers tend to apply the term to a variety of processes such as automation, machine learning, and optimization. But that has led to a muddled understanding of the technology. And martech vendors aren’t exactly taking steps to clear up the confusion. As AI takes its place as the industry’s buzzword du jour, companies everywhere are hoping to ride the hype.
Henry Schuck, CEO of DiscoverOrg, a sales and marketing intelligence solution, believes that Salesforce’s Einstein announcement in 2016 was mostly a rebranding of what the company formally called “predictive lead scoring.” According to him, the new terminology has led to a hysteria of “rebranding” from companies desperate to keep up.
Sameer Patel, CEO of the B2C smart automation software company Kahuna, worries this widespread adoption could have real consequences. “There’s just a lot taking old technology, plastering AI on it, putting a nice gloss on it, and saying, ‘Hey, great, I’m AI,'” he said. “No, you’re not. It’s going to give AI a bad name, ultimately, because the results won’t come out.”
Unfortunately, there is no authority that defines what exactly is and isn’t AI, and companies will likely use the term’s loose definition to hype their own products or undercut competitors. AI’s definition will always be broad—what really matters is what the technology can do.
The rise of AI
Jon Epstein—CMO at Sentient, one of the most-funded independent AI firms in the world—believes AI will be more revolutionary than the internet.
But artificial intelligence has been around for years. The difference is that recent technological innovations have propelled its rise to the front of the marketing hype train. As Scott Brinker, founding editor of ChiefMartec, wrote last year, “The horsepower and data to efficiently run AI algorithms is now within reach of everyone.”
AI is already visible in a number of marketing disciplines. Ad tech companies use AI to power programmatic ad exchanges, automatically optimizing ad campaigns that once required intense menial labor. Marketing automation software, such as Marketo, also takes care of tasks that once required hours of labor, such as sending nurture emails or qualifying leads.
“Any part of the marketing world where a marketer has to read data and make decisions based on that data will be affected by AI in one way or another in the near future,” Schuck said.
“With any AI, they are really only as good as the data that goes into them.”
While machine learning and automation tools can simplify marketing processes, many of today’s AI companies are also focused on how technology can improve customer experience. Sentient, for example, uses its “distributed AI platform,” which runs on an incredible network of 2 million CPUs and 5,000 GPUs (the horsepower in a computer), to help retail clients optimize the customer experience. One of its two marketing products, Sentient Ascend, does A/B testing in real time at massive scale.
“Normally when you’re testing … it’s pretty slow and painstaking.” Epstein explained. “Instead, using a different type of algorithm, you can test ten, twenty, thirty changes at a time, and it solves not only what changes to make, but what combination of them to achieve the highest level of response.”
To visualize this, think of testing colors and fonts on a lead form, the centerpiece of many marketing tool kits. Epstein told me that after one company used Sentient to test 380,000 form designs, its lead form conversion rate increased by 45 percent.
“The fact that you can pick up that kind of gain from what you would think of as tweaks, it’s pretty profound,” Epstein said. “We think there’s a lot of money being left on the table.”
At Kahuna, which serves mostly B2C clients, Patel values AI for the way it can help customize customer experience, using machine learning and signals to optimize for individual needs. Some companies are already showing that this customer-focused approach works.
“If you look at how quickly you can call for an Uber or pick a movie on Netflix, our tolerance and our attention span right now as consumers is wildly, wildly, wildly shorter than it used to be,” Patel said. “You’re dealing with a world where they’re expecting that you know who they are, you can understand their needs, and you can fulfill them in real time.”
From teen to neo-Nazi
While it’s easy to get caught up in the hype, AI still comes with issues. These software products rely on integrated and accurate data collection systems, but not every company has the digital infrastructure in place to take advantage of such advanced technology. Since martech is a young space, many marketers still use poorly integrated systems that lack oversight.
“With any AI, they are really only as good as the data that goes into them,” Schuck said.
In 2016, Microsoft built a Twitter bot called Tay meant to resemble a teenage girl that could communicate with users. In less than 24 hours, the bot tweeted messages like “Hitler was right I hate the Jews,” and “I fucking hate feminists and they should all die and burn in hell.”
Since the bot was programmed to take signals from users, internet trolls overwhelmed it with offensive messages and quickly turned Tay into a neo-Nazi bot. The controversy was an embarrassing example of AI’s potential to spiral out of control.
AI also dregs up criticisms often leveled at the adtech industry: namely, the reliance on over-personalization. Privacy concerns about too much targeting could lead to an “uncanny valley,” where customers are creeped out by AI’s abilities to predict their wants and needs.
However, most marketers remain confident the value outweighs these concerns. “I feel like we’re getting to a place where, as long you’re not messing with the user’s trust, people want personalization,” Patel said. “I think we’re moving from being okay with that to actually expecting that.”
The ghost in the machine
So here’s the big question: Where does artificial intelligence leave the marketer? Technology has already reshaped blue-collar industries, eliminating the need for jobs. Won’t automated systems threaten to destroy traditional marketing roles as well?
AI proponents concede that technology will replace some roles, but they believe AI has the potential to create just as many new jobs. Plus, they argue, software could free up marketers to focus on creative, strategic work rather than day-to-day processes.
“Ninety-five percent of the people who came to marketing did not come to this data-driven world. They came for the soft side of marketing. ” Patel said. “Because of AI, you can be that creative person you always wanted to be. You now have systems that don’t require you to become a mathematician.”
“Because of AI you can be that creative person that you always wanted to be.”
Perhaps there’s a compromise in the future. Patel and Schuck both expect a new breed of marketer to emerge—someone part data scientist, part creative, who can work directly with AI systems.
“AI will significantly simplify the user experience for many of the martech tools that marketers use, as well as automate a tremendous amount of the ‘manual’ labor associated with marketing programs today,” Brinker added. “It’s unlikely that we’ll be lounging around while robots feed us grapes.”
Currently, Sentient is working on building an “AI that writes AI,” a holy grail that could exponentially accelerate the sophistication of existing software. What happens if a competitor gets its hands on such technology while your company cannot? As AI develops, marketers will have to adapt quickly, and some of them may not be prepared.
“We’re at the very beginning of something,” Epstein said. “Seeing what our own technology can do, the stuff we’re doing research on—we ain’t seen anything yet.”Image by Zack McDowell