Content Performs Better with AI
Content marketing is kind of like coaching a team sport. It takes a great game plan plus high-performing assets to win. And, like a team, your content strategy is only as strong as its weakest link. You need to be able to understand the strengths and weaknesses of each asset in order to field the strongest possible roster.
Traditional analytics can provide you with a lot of data about the effectiveness of each asset, so it’s easy to spot high and low performers. But it’s not always easy to understand why a content experience performs well or does poorly. Understanding that often takes professional analytical analysis, or—worst case—relies on a marketer’s intuition.
“Normal analytics are typically page based. When a marketer logs into their analytics solution they’ll see numbers and a set of key performance indicators (KPIs). So, while it’s easy to see that a particular page is associated with a high or low KPI, it’s really up to the marketer to figure out why,” explains Jonas Dahl, product manager for machine learning in Adobe Experience Manager.
Jonas continues, “It’s not always easy. You might see that theme “A” did well, and theme “B” did poorly and assume that you need more of theme “A”—but what if the true difference was due to the author or the quality of an associated photo? We wanted to see if we could use AI to get past bias and intuition, go deeper into a site’s data, and add meaningful analysis to our traditional insights.”
The result is game-changing sneak technology called Intelligent Content Insights. Recently demonstrated at Adobe Summit, it can leverage your website analytics to identify low performing content, compare its qualities against content from your past campaigns, and make targeted recommendations based on predictive performance.
Intelligent Content Insights works by using AI to analyze and understand your site’s content at scale. By using features powered by Adobe Sensei, it can scan hundreds, or even thousands of different content experiences, performing image and object recognition on videos and photos as well as text, keyword and sentiment analysis to generate a set of rich metadata about your content.
“We need to generate a similar understanding to the human interpretation of content to use as the foundation of our analysis: What’s the topic? What’s the sentiment? What’s in the photos and videos? We can leverage all of the different Sensei technologies to generate that understanding, and then compare it to the actual performance of content assets over time using data from Adobe Analytics,” says Jonas.
The combination of a deep understanding of the site content at scale, along with traditional metadata opens up new possibilities for insight. In his demonstration, Jonas showed how a marketer can select between different audience segments and KPIs to generate a ranked list of content elements that have the most impact on a particular KPI. You might find, for example, that videos have the highest impact on engagement for your site. It can then highlight which specific content assets conform or diverge from the characteristics of your best-performing content.
Although Intelligent Content Insights is still just an early look at a developing technology, it provides a tantalizing glimpse of the kinds of capabilities Adobe Sensei is adding to the marketer’s toolkit.
“One of the things I love about being at Adobe is that we do a lot of broad research, but our goals are centered around creativity and marketing. It makes a lot of the things we work on converge. We can develop something for one use case, but then build on top of it for another to advance the research really quickly,” says Jonas. “Using data to extract meaning about which experiences work and which don’t has traditionally been one of the hardest things to do in analytics, so it’s really fascinating to see the impact AI is having on our ability to get a deeper understanding of content and its impact on the customer journey.”
This story is part of a series that will give you a closer look at the people and technology that were showcased as part of Adobe Sneaks. Read other Peek Behind the Sneaks stories here.