Content is king, say marketers. What they remain mum about: The data that informs this content is now a dominion so vast that no monarch may ever get it under control. Savvy leaders are responding by exploring how artificial intelligence (AI) might sustain the high-speed evolution of the entire industry.
With precisely targeted marketing as their guiding light, content producers are examining technology that can organize data to create the next successful piece of content. That’s because the inexhaustible amount of data generated by e-commerce offers a good foundation for the making of content, but there’s so much of it that marketers’ efforts can’t be achieved by humans alone. As marketers tweak how they measure the efficacy of their content, they’re also letting machine learning take on aspects of its creation.
To be clear, marketers aren’t being replaced — the age of the robot ad executive hasn’t yet dawned — but the industry knows where to look for help. Each iteration of various AI applications provide marketers with opportunities for increasingly sophisticated analysis and the chance to experiment with adventurous new ways to connect content with audiences.
Liesl Leary Perez, head of global content marketing at software and content management giant SDL, says the struggle is only getting tougher. “We all know we are drowning in information and are inundated on a level that’s impossible to even consume,” she says. “The core mission of many businesses and organizations was producing information [or what we call content today], and now we have so much that we can’t even make sense of it all.”
The age of the robot ad executive hasn’t yet dawned.
That’s particularly problematic for marketers and advertisers, says Ali Haeri, vice president of marketing at Steelhouse, builder of adtech platform software. “Advertising is going through something like the transformation Wall Street went through with the advent of high-frequency trading,” he says, referring to the 2005–2009 leap in algorithmically determined trading strategies that altered how financial exchanges traded everything from stocks to complex financial derivatives.
“Marketers are now dealing with these gigantic data sets, and until now, they’ve had to do everything manually, which was clunky and didn’t give them the ability to deal with people on a one-to-one basis. AI allows us to create one-to-one experiences that are more efficient, and every marketing touchpoint is going to be able to be serviced by AI,” Haeri continues.
Discussions with marketers, data scientists, academics, and programmers reveal the increasingly close relationship between AI and the future of content, as well as the directions it might be headed in each of their respective fields. While the challenge of creating “smart content” for consumers might differ between marketers trying to sell products and support brand identities, there’s broad agreement that AI is playing a vital role in these processes.
How we got here
For decades, retailers, marketing departments, and advertisers struggled to keep track of what’s now called “the customer journey,” the roadmap that addresses the entirety of a customer’s awareness of and interactions with a product, service, or brand. As data generated by e-commerce grew in volume, so did its value in tracking online customers. With e-commerce sales representing 10.7 percent of all retail transactions in the second quarter of 2019 — an estimated $146 billion, according to the Department of Commerce — the data explosion means there’s more information available that can be organized and refined into effective content. Effective personalization can cut customer acquisition costs by as much as 50 percent and boost revenues as much as 15 percent, according to a Harvard Business School study, so it’s well worth the effort.
Consumers’ use of social media and embrace of e-commerce creates a wealth of data for marketers, and as methods for learning from that data get more sophisticated, the uses for that information (and the methods employed to make marketing efforts more effective) become, for lack of a better word, smarter, says Janet Polyakov, vice president of product marketing at Terminus, a San Francisco account-based marketing firm. “Calling it ‘smart content’ is sort of using a new buzzword to describe how data is used,” she says. “What that really means is that we use the data tools that are available to us to find your ideal customer profile.”
One of the first ways this process was honed was via natural language processing (NLP), which, among other applications, can be used to help companies’ call centers by employing a text-mining tool to discover trending topics. According to William Fehlman, the former director of data science at USAA, this “topic modeling” helped analysts improve the customer experience. In a July 2019 podcast of This Week in Machine Learning & AI, he said, “We created topic models that had categories of terms that let our analysts look at them and say, ‘I understand what this is about.’ The only limitation is the processing speed we’re working with.”
Current advances in AI applications, such as SDL’s latest program to tag, translate, and summarize content for chatbots serving a global customer base, demonstrate the leaps being made in how data can be sorted, synthesized, and processed to become usable marketing content, says Leary Perez.
“There’s almost too much information even for machines to consume now, so AI is coming into our world to help us decide [what data is effective],” she says.
However, the goalposts that determine which data is useful are constantly moving. Companies aiming to connect to an individual consumer have had to adjust to a decade of shifting parameters for effective marketing and sales as they seek to incorporate AI into their workflow. Leary cites a Forrester Research study commissioned by SDL last year that shows a direct link between better customer experiences, revenue growth, and lower costs. “Since customers and prospects have raised their expectations when assessing a product or service provider,” the study notes, “companies must rethink the content that customers consume across their buying journey.”
This challenge will become easier for AI as older, more privacy-conscious consumers are increasingly outnumbered by digital natives who have fewer qualms about providing the data needed for this personalization. Indeed, the more personalization millennial shoppers get, the less concerned they are with who has their data, according to a 2018 Salesforce report.
Not all data mining leads to gold
As this tilt toward personalized marketing continues, though, critics question the balance of power. Dr. Robert Epstein, a leading behavioral psychology expert now at the University of California San Diego and veteran commentator on the uses of digital data, is quick to see a downside. This summer, he warned the Senate Judiciary Committee of Google’s potential power over voter behavior in the upcoming election cycle and says the ability to manipulate data has hidden, heavy costs.
“You can’t do customization without massive surveillance, and the aggressiveness with which tech companies are pursing this data is, in itself, a problem,” he says. “When your phone is outputting 11.4 megabytes of data a day, even if you wanted to defend this based on consumer convenience, it’s completely out of hand.”
Leary Perez sees the stakes for companies as existential and believes younger consumers will accept a little more lost privacy for the gain of increased personalization. “If they don’t have a good customer experience, they will drop off,” she says. “The organizations that are going to continue to be successful are the ones that really fulfill this dream of a frictionless customer journey.”
Using AI to contain endless possibilities
As companies begin to understand both the value and importance of their content (and how some of that data may have been procured at the cost of customer privacy), the best of them recognize the vital need for a carefully secured but potent cohesive brand experience for every customer, says Leary Perez. Smart uses of AI can create and implement holistic content management strategies to ensure every piece of product, marketing, sales, and support information remains up-to-date and personalized for real-time customer needs.
Getting these in place will free creative to make better decisions themselves, says Haeri. “Marketing has become data driven, but AI is definitely going to take the wheel in terms of the ongoing maintenance of these campaigns,” he says, though it will still be up to creatives to generate ideas. “The creative aspect of brand marketing will not be touched. There’d have to be a pretty seismic shift in AI where it could convey a brand voice and mission and really evoke some sort of emotional response.”
So, while AI-supported content isn’t replacing humans — yet — marketers know it’s given them increased abilities and multiplied the power of their content beyond a human scale. Productivity and versatility from AI applications now outpace everything but the creative, artistic side of content creation. And in an age where content is king, content’s strength now comes not primarily from the people who create it, but from the vast amounts of data used by current and future AI applications. That’s what’s really holding up the throne.