Wharton Research Looks at The Future of “Content Engineering” for Marketing

When most people hear the word “engineering,” they think of urban infrastructure or electrical switchboards, but now some of the world’s top business professors are throwing “content engineering” into the mix. This expansion of vocabulary might sound like the latest buzz phrase, but it actually represents a material advancement for one unlikely area of creativity: marketing copywriting.  

What’s content engineering, and what did the research prove?

Professors Kartik Hosanagar of Wharton, Dokyun Lee of Carnegie Mellon, and Harikesh Nair of Stanford define content engineering in their recently published research as the data-informed design process of marketing language that optimizes consumer engagement. The study, “Advertising Content and Consumer Engagement on Social Media: Evidence from Facebook”, used natural language processing technology to examine +100,000 Facebook posts in order to determine which brands best engaged their customers. They found that emotion drove over a 20% increase in a message’s number of Likes. For example, adding emotional content adds about seventy more Likes and five more Comments on average, all things held equal.

What’s next? Implications & applications

The intersection of marketing and technology is not new, but its scale and scope certainly are. In an interview for the Knowledge@Wharton podcast, Professor Hosanagar said, “Next, we want to take those techniques and apply them to large-scale analysis of content broadly…. We hope to go there. It’s a matter of getting lucky again with a good data partner.”


Through our content generation solution, Persado has essentially been performing the same analysis as Drs. Hosanagar, Lee, and Nair for several years now. Using natural language processing, natural language generation, and advanced predictive analytics, Persado’s AI platform provides tagged and scored content for tens of millions of marketing messages on email, Facebook, SMS, display, mobile push (and more), that has amassed over one billion impressions and counting, giving us granular insight into content performance. One thing we learned? Emotions contribute as much as 60% to response variance.

Professors Hosanagar, Lee, and Nair–we’re ready when you’re ready.

Dr. Kartik Hosanagar, Penn’s Wharton School of Business

Dr. Dokyun Lee, Carnegie Mellon’s Tepper School of Business

Dr. Harikesh Nair, Stanford Graduate School of Business