Content is king, but that doesn’t matter much if nobody is reading it. 

Consumers are deluged with so much information from so many different sources – there are more than 4.5 million daily blog posts alone — that it’s hard for marketers to know whether their messages are getting through. Clicks and page views can be impressive metrics, but they don’t indicate whether users actually read and absorb the content, or whether it directly motivates them to make a purchase. While recent studies have focused on different angles of audience engagement, there’s new research that examines exactly what makes people go beyond the first click and dive deeper into the content. 

Jonah Berger, marketing professor at The Wharton School of the University of Pennsylvania; Wendy W. Moe, marketing professor at the University of Maryland’s Robert H. Smith School of Business; and David A. Schweidel, marketing professor at Emory University’s Goizueta Business School are co-authors of the study, “What Leads to Longer Reads: Psychological Drivers of Reading Online Content.” The scholars used natural language processing to analyze data from more than 700,000 reading sessions from over 35,000 articles from nine major online publishers. They combined that data with three experiments to determine how emotion and word choice shape reading behavior.

“If content creators want people to read their content, they have to understand what drives sustained attention in the first place,” the professors wrote in the paper. “This work provides a set of actionable directions … that creators can use to craft content that encourages reading.”

MSI spoke with Berger and Schweidel about their study. Their answers appear below: 

MSI: You wanted to uncover what makes people want to keep reading online. What did you find? 

Jonah Berger: Lots of individuals and organizations want people to read their content. From newspapers to politicians to nonprofits and for-profits, everyone wants to create content to encourage people to pay attention and eventually become customers. But that content’s not going to be very useful if people don’t read it. So, what leads to longer reads? We found a couple of things. First, we found that emotion plays an important role, but it’s not just that any emotion increases reading. It’s the associations with particular emotions beyond positive or negative. Some emotions make us feel good, some make us feel bad. 

Emotions also differ on a number of dimensions. One, for example, is arousal. Anger and anxiety are high-arousal emotions. They fire us up and they drive us to take action, whereas sadness is a low-arousal emotion. It deactivates us and powers us down. The same thing can also be said for dimensions like uncertainty. Some emotions are associated with more certainty. When we’re angry, we feel very certain. Other emotions, like anxiety or fear, are associated with uncertainty. The results show it’s not just that emotions increase reading, but it’s the tendencies that go with those emotions. Emotions that make us feel uncertain, like anxiety, increase the likelihood that we read because we want to figure out the answers, as well as emotions associated with arousal that fire us up, that drive us to take action. Continuing reading is one of those actions. 

While those ideas might be a little bit complicated, I think the key insight is it’s not just that emotion increases reading, and it’s not just that positive emotion increases reading or negative emotion decreases reading. It’s about the emotions and the activation tendencies associated with those emotions. 

David Schweidel: We find that emotions help as far as engaging with the audience and getting them to continue reading. We also find that the easier it is for the audience to read, the more likely they are to keep going. More complex and abstract language decreases ease of processing and the chances that someone keeps reading.

Berger: Abstract concepts like truth, honor, kindness and grace are intangible things that are hard to understand. Cats, dogs, deep spruce green – all those are examples of things that are concrete. We could use more familiar words to refer to the same things, rather than less familiar words. I think this is really important because you might think, as a content creator, I have to change what I write about to get people to read it. It’s easy to get people to read about sports, for example, or celebrities, but it’s hard to get them to read about environmental policy or tax reform or B2B services. There are certain topics that people are more likely to read about and less likely to read about…. But what our results also show is that even when people are talking about the same topic, we can use different language to make them more likely to read about it. So, the language we use can have an impact on consumption above and beyond what specifically we’re talking about.

MSI: Does that mean that every message needs to be dumbed down or to pull on the heartstrings to be effective?

Berger: I want to be very careful. We’re not at all saying things need to be dumbed down; that’s the exact opposite of what we’re saying. We’re saying you need to use types of language that are easier for people to understand and parse. The same thing goes for pulling on the heartstrings. We’re not saying, ‘Oh, you just need to use emotions to get people to read.’ There are certain types of emotions that will increase reading, but there are other types of emotions that might even decrease reading. And it goes even deeper than emotion: We can use language that is certain or uncertain, regardless of whether it’s emotional or not. It’s not about emotion, per se, it’s about the things that underlie that emotion.

Schweidel: There are two components to stories: the substantive content, and then the way in which it’s communicated. Yes, the simpler you can make it, the more likely readers are to continue. But some ideas are inherently complex, like public policy. We’re not saying to shy away from the complex topics, but that the language used throughout the article matters. You can talk about public policy in an abstract way, or you can make it as relatable and digestible as possible. We’re suggesting that writers and publishers do the latter.

MSI: Your study focused on the consumption of news, sports and feature articles, which are quite different from display advertising, commercials and promotional emails. How can marketers translate the results of this study to start creating more meaningful content for their target audiences? 

Schweidel: In marketing, we focus a lot on size. How much traffic are we attracting? How big is the audience that we’re reaching? But what we don’t discuss nearly as much is the quality of the audience that we attract. For marketing content to be effective, a few things need to happen. We need to reach the consumers, which requires cutting through all the clutter that bombards them. Once we get consumers’ attention, we have to engage them. We know how to create headlines and content that people will click on. If the intent is to get that consumer to subscribe to a publisher’s content or to persuade a website visitor, they need to go beyond that click and actually consume the content.

Berger: We looked at a wide variety of information. Yes, we looked at things like news, sports and feature articles, but also political news, technology information, all sorts of different things. Companies all the time are trying to get people to read their stuff — whether I create a white paper that I want people to read, whether I create a blog post that I want people to read. As marketers, we need to understand the language that leads to longer reads. How can we write things in a way that makes it more engaging? We may not be able to change the topic of what we are writing, but by changing the way we write about it, it’s going to be much more impactful.

MSI: In your paper, you mention the common refrain that it’s easier to get people to pay attention to lightweight topics, like celebrity gossip, than more serious matters, like climate change. However, you say there is hope for organizations that are trying to generate attention for more complex topics. Could you explain that?

Berger: It’s not just what you say, it’s how you say it. I can write about climate change using familiar language or unfamiliar language. I can talk about climate change by saying the Earth is getting warmer, or I can say that ice is melting. “Ice is melting” is very concrete language. It’s easy for people to understand the idea of “ice is melting.” They can visualize that. Global warming is much more concrete; we get a sense that temperature is warming. Climate change is pretty hard to understand, so even when talking about a topic like climate change, using concrete language – polar ice caps are melting – is a very concrete way to communicate the change that is occurring.

Schweidel: Publishers have different audiences that they are trying to reach, so most are not going to consider dramatically altering the topics that they choose to cover. But even for the most dense topics, the way in which it is communicated to the audience makes a difference in how much of an article will be read. More concrete and simpler language makes it easier for readers, and that will translate to increased reading. They may not read as much on dense topics as they would of lightweight topics, but the choice of language used to discuss a dense topic will certainly affect how much they choose to read.

MSI: What’s next for this line of research?

Schweidel: Textual content offers a lot of opportunities for exploration. While we focused on the content here, there are also factors relating to the audience that can be further investigated. For example, does it matter how diverse the audience is? To what extent should content be tailored to the audience? We know that social effects exist in that articles are shared, but are there such effects for how much of the article is read? 

More broadly, I’m interested in the flow of information, where content and social network structure both play an important role. Going beyond text, we need to further our knowledge of images and videos via analytics. Ultimately, I think this will lead to work in automated content creation. That is, how good a job can a machine do at creating content, and what are its limits?

Berger: I think this is a really interesting variable that no one in marketing has really looked at before. Marketers have been interested in sharing and clicks, but no one’s ever looked at actual reading, actual engagement with written content. So, I think this is a really exciting window into a variable that’s going to be more important as we move forward. 

If you look at a lot of the online metrics for content, there are things like views and shares. Views are a really bad measure of content. All they say is that someone clicked on it, but you have no idea whether they actually engaged with it. Same thing on social media. People will pay a lot of attention to friends and followers, which is a really bad measure. It doesn’t tell you whether people engage with the content in any way. When we look at things like reading depth or dwell time – how long you spend watching a video – those are much more important metrics that are more likely to be linked to the outcomes that we care about. We don’t want people just to click on our content and go away to something else. We want them to stay and pay attention to it, and this research is starting to provide insight into that question.