Why the Medium Shapes the Message in Marketing

Consumers love to share their opinions, whether it’s a scathing one-star review on a hated product or telling friends about a delicious new restaurant. There is plenty of marketing research on what consumers love to talk about, but far less on how they talk about it. Is talking better than writing to get the message across, or is writing more influential?

Wharton marketing professor Jonah Berger explores these different ways of communicating in his latest research paper, “Expression Modalities: How Speaking Versus Writing Shapes Word of Mouth.” It was published in the Journal of Consumer Research and co-authored by Matthew Rocklage, marketing professor at Northeastern University, and Grant Packard, marketing professor at York University in Toronto.

Berger spoke with Knowledge at Wharton about the paper, which is part of his continuing research into the role of language and communication in marketing. Listen to the podcast above or keep reading for an edited transcript of the conversation.

Knowledge at Wharton: Let’s start with an overview of this paper. What were you and your co-authors trying to ascertain?

Jonah Berger: I don’t have to tell you that companies, consumers, and other marketplace actors are constantly communicating. If we think about consumers, for example, we share word of mouth about products we like, services we don’t like, experiences that we’ve had. Companies communicate with consumers through customer service and other means. Financial service agents talk to their clients or prospects. Doctors talk to their patients. Politicians talk to prospective voters. A range of marketplace actors is constantly communicating with various audiences in one way or another.

But while we think a lot about what we want to say, we don’t often think a lot about the medium or modality we communicate through. Sometimes we communicate through voices, sometimes we use writing. We text someone, we send an email, we write a presentation that they may read at a later date. The question, we wondered, is whether this seemingly subtle shift — speaking versus writing — might shape what we communicate. Whether the medium we communicate through might shape the message.

As a consumer, think about communicating with someone. Maybe you have a product that you love or hate, and you want to tell someone about it. Would calling them on the phone versus writing them an email lead you to share different things about that experience?

Or take a company. Many companies right now are thinking about whether to shift customer service from a traditionally spoken means — usually you call customer service — to a written one. You text, either through your computer or phone, with service agents. Might that shape what the consumer or agents communicate, and how might that affect how that interaction goes?

These are the types of questions we were interested in studying. Does it change the language we produce, but also the consequences of that language or downstream outcomes of that interaction?

Knowledge at Wharton: You conducted six studies for this paper. Can you tell us about them and the results?

Berger: I’m going to skip over a few of them and just get to the main ideas. In our first study, we just wanted to see, does this happen in the real world? If we go out there and look at things that people write and things that people share through voice, do they differ? Do they differ in a particular way we might expect? And what might the consequences of that be?

We got thousands of written reviews from Amazon.com, and we got thousands of spoken reviews from YouTube, and we analyzed the language contained in them. In the written reviews, you obviously see the language already. We transcribed the spoken reviews to get the words contained in there. And importantly, we matched the products. If we looked at a certain type of product on Amazon and a completely different type of product on YouTube, it wouldn’t be surprising if people used different language. But we picked categories where we could find the same exact product being talked about on Amazon and YouTube to see whether the way that people talked about it differed based on the modality they communicated through.

What we found is that written reviews were much less emotional. What do I mean by that? Well, they use less emotional language than spoken ones. They use less highly emotional words and use a little bit more cognitive language explaining what something does or describing it even in positive terms. That difference had an important impact. We found that reviews that use less emotion were, in fact, less impactful. People found them less helpful and less useful.

Now, you can see that study and say, “Oh, that’s interesting, but there are so many differences between Amazon and YouTube, between different people writing different reviews. How do you know it’s really about speaking versus writing that’s changing things?” So, we conducted a simple experiment where we held everything else constant. We took a set of people. We asked them what they thought about a restaurant, and we either asked them to speak about their thoughts about that restaurant or write about it. Talk into a microphone as you would if you were calling a friend, for example, and telling them about your experience. Or write down that experience, as if you were texting or emailing your friend about it.

Even though it was the same exact people randomly assigned to these two conditions, merely changing how they communicated changed what they talked about. We found the same thing where speaking was much more emotional and writing was less emotional. Writing involves more deliberation or thinking about what to say, and that makes what we share less emotional.

Knowledge at Wharton: Deliberation is a word that kept coming up in your paper that really caught my attention. What is the key here with deliberation?

Berger: Yes, that’s one thing I find really fascinating about this paper. Again, when we are producing language, when we are producing content, we don’t think a lot about how the mode we’re communicating through — speaking versus writing — is changing that content. But think for a moment about this interaction that we’re having. You asked me a question, and I, trying to be a good interviewee, took a few notes beforehand. I have some notes on the page in front of me, but I’m still responding on the fly, figuring out exactly what words to use.

In most spoken interactions, we don’t even have notes ahead of time. We’re thinking about what to say and producing it at the same time. We don’t have a lot of time to think about what to say or how to organize it. Contrast that with writing. When someone asks you something over written communication, or you respond to them using written communication, you have a lot of time to think about and construct what to say. You can edit it, reorganize things. You have a lot more time to deliberate before producing that content. Those differences between those two modalities change what we end up producing, not just in terms of emotional language but in other types of language as well. The means we communicate ideas through actually change what we end up communicating by the nature of those mediums.

You can even think about the same idea in terms of negotiating. The more notes and things you write down ahead of time, the more organized you can be in your thoughts. Not only are you less emotional when you write, but you can do a number of different things to organize the content that you end up producing. Nowadays when we think about content, we think about online content and social media posts. But when you’re speaking to your boss, you’re producing content. When you’re a financial service agent talking to a prospect, you are producing content. You are producing language that can impact the audience you are communicating with. And the mode we communicate it through can shape that discussion.

Knowledge at Wharton: What does this study contribute to the field?

Berger: I think this is one of an emerging set of papers that is beginning to look at how the medium shapes the message. The advent of the typewriter or the computer made it easier to have written communication. Most recently, text messages made it much easier to shoot off quick missives to other people, and now even companies use those to interact with clients.

Some of us may have begun to experiment with voice-to-text. The content is delivered to the audience through text, but we are speaking to record it. You may shoot off a test message, not by writing it with your fingers but by speaking into your iPhone because it’s easier. You may write an email using the dictation function in Microsoft Office.

Similarly, lots of companies and organizations are thinking, “Great, if I could choose speaking versus writing, or texting versus email, how should I communicate with the audiences I want to communicate with?”

This paper is helping us understand how the mediums can shape the message, how the way we produce this content can shape the content that ends up being produced.

Knowledge at Wharton: I’m curious which one you think is better — writing or speaking?

Berger: The answer is, “It depends.” It’s not that speaking is better than writing, and it’s not that writing is better than speaking. It depends on what you’re trying to achieve with that interaction. If you want to be more careful and reasoned, writing is pretty good. It gives you the time to construct and find what you’re going to say. On the other hand, we have a lot of data in this paper that suggests that emotional content is often more impactful in a positive way. So, if you want to be impactful, speaking can be good to be persuasive to change others’ minds.

On a website, do I want to use spoken or written reviews? Well, that choice is going to impact not only what kind of content people produce, but how likely it is that that content is going to be persuasive. If I’m a brand, for example, and I’m encouraging people to create product reviews, it might be better to get them to speak because they will be more emotional. And in many product categories, that might be more persuasive.

Similarly, we talked about customer service. The mode of interaction may shape the nature of that interaction. If I’m a doctor or a lawyer, you could say, “I want to reason through my arguments first. I want to write them down.” But if I want to be particularly persuasive, maybe I need to be sure that’s not sucking out all of the emotion because that may make it feel lifeless when I communicate it.

I think it’s really understanding the impact we want our communications to have, and then taking a step back and analyzing the language itself. I think there are a lot of interesting questions here to be answered in future research, but the main idea is understanding the interaction between the medium and the message.