As the volume of digital content grows exponentially, marketers are spending more time than ever trying to understand what resonates with their customers. Over the next couple of years, artificial intelligence (AI) could boost these efforts by pinpointing the best emotional appeal, subject matter, style, tone and sentiment to focus on.
U.S. marketers spent more than $10 billion on content in 2016, according to a Forrester estimate. For many marketers, their content marketing strategy now includes a wide range of tactics like social media posts, short- and long-form video, live video and sponsored articles. However, for most, content is delivered with a piecemeal approach. A different Forrester report titled “The Top Emerging Technologies For B2C Marketers” suggests AI-based content intelligence solutions hold the potential to provide marketers with a holistic, automated approach to creating, managing and optimizing content — a strategy it refers to as "content intelligence."
While only a few content intelligence solutions exist today, Forrester expects more to become available and for the technology to have a disruptive impact within one to three years. By ingesting unstructured content, assessing it against a reference set and analyzing the content, the technology is able to understand and capture the inherent qualities of content and delivering actionable insights.
“The intersection of content and cognitive is also a big area of opportunity — for both advertising and marketing campaigns — particularly creating more effective content, creating content variations, dynamically assembling content, helping figure out what types of content and how much content needs to be created,” Maria Winans, CMO IBM Commerce, told Marketing Dive.
Understanding context
An example of content intelligence in practice is the IBM Watson Content Hub which uses cognitive capabilities to understand and learn about the data in a company’s content management system in order to automatically tag the image, video and document content based on millions of previous examples.
IBM recently announced an upcoming video enrichment service that will tap a number of Watson APIs including Tone Analyzer, Personality Insights, Natural Language Understanding and Visual Recognition to generate video content insights with even deeper understanding of context and content than currently available.
Google is also testing its Google Cloud Video Intelligence API designed to use object recognition to allow clients to automatically classify the content of videos. However, the technology hasn’t been perfected just yet as the University of Washington was able to trick the tech’s ability to correctly tag video content.
These efforts to put more technology behind content strategies come at an interesting time as some brands are failing to hit the right emotional tone during a period when consumers emotions are running strong, begging the question can AI actually do a better job of making an emotional connection than human marketers can. The AI push also coincides with a bigger focus on leveraging the expertise of writers and producers with experience in the entertainment industry, like Dove's recent hiring of Shonda Rhimes, who is behind hit TV shows such as Grey's Anatomy and Scandal, as they look to make an emotional connection.
Where technology and application intersect
In practice, marketers are not running out to invest in AI technology. Still, many are thinking about how their content performs in campaigns and resonates with audiences.
Where the technology-driven definition and solution meets customer experience is where marketers may find the true value in content intelligence. By taking a holistic view that includes both better understanding the growing portfolio of content assets marketers are accumulating a la Forrester’s definition as well as the insights, measurement and key performance metrics around content marketing campaigns, brands can start to make sense of how to boost their content efforts.
When marketers know what works with an audience, and also know attributes about their content such as topic, sentiment and emotional appeal that is tagged and easy to find, creating highly targeted and personalized messaging become much easier.
“Content intelligence emanates from the underlying data tied to marketers’ content marketing initiatives and provides them with the insights to understand what’s working and why so they can make informed decisions about where to invest their content marketing dollars,” Steve Sachs, CEO of cross-channel content personalization platform OneSpot, told Marketing Dive.
The value is in understanding down to an individual what content is resonating in order to create a deeper level of engagement with customers and drive measurable return on personalization, said Sachs, calling the metric ROP.
“Using AI and natural language processing to gain a richer understanding of their audience’s interests and automatically match the right content to the right person — anywhere that person may be online — has tremendous value," he added.
For understanding audiences, Sachs said marketers should strive to track and analyze engagement down to specific topics and subtopics for granular details on that content engagement.
“For a food brand customer, for example, we can see which recipes and specific ingredients are receiving the most engagement, and that may vary depending on which parts of the country we’re analyzing," Sachs explained. "This can help a marketer decide to develop and deliver more recipe content featuring the popular ingredient or change out a less popular ingredient.”
Putting content intelligence into action could also include using those insights to inform editorial strategy. If that food brand marketer found that content about garlic drew a large, highly engaged audience but relative to the website’s full portfolio of content there wasn’t a lot of content about garlic, the marketer just uncovered a strategic opportunity to create more content about garlic to drive deeper engagement.
How bots complement content intelligence
On the campaign data side of content intelligence, Kate Richling, General Manager and Head of Marketing at Rova, said marketers need to think past metrics like impressions, clickthroughs and even content recall and user expectations to include business KPIs like new leads, customer acquisition cost and net promoter score.
“When you’re able to align your content intelligence efforts with other tactics your organization is executing, connecting data points to ladder them up to business outcomes, that’s where the real ‘intelligence’ comes in,” Richling told Marketing Dive.
Where marketers can gain bottom-line results is making use of the business metrics to focus on what a content marketing program is really trying to achieve. As Richling explained the most active and engaged audience might not be a business’ actual base of potential customers if those ROI-oriented results aren’t strong.
She also sees another area of marketing artificial intelligence as a place where marketers can gain insights to improve content marketing.
“I think there’s some really interesting things happening with customer communication tools that can work complementary to content intelligence,” said Richling. “For instance, a potential lead visits a website — a bot chats with them and asks a question to prompt deeper engagement. They interact and converse — all without a human being on the brand side through the use of advanced bots and automated responses. This can help brands make their content more impactful, learn more about their readers and identify what’s working and what’s not beyond just impressions.”
Content intelligence really boils down to personalization according to Sachs, and he anticipates the future of the tech will allow for greater level of one-to-one personalization with the benefit of easier, less rigid and less costly implementation.
“Instead of ‘set it and forget it’ approaches to marketing orchestration, segmentation and other marketing tactics, artificial intelligence — when applied properly — will make marketing experiences richer, smarter and ultimately more effective,” said Sachs.
