What to Say to Get Your Way | Jonah Berger
48:14 Share

Welcome to the Talks at Google podcast, where great minds meet. I'm Abhay, bringing you this episode with professor and author Jonah Berger. Talks at Google brings the world's most influential thinkers, creators, makers, and doers all to one place. You can watch every episode at youtube.com slash talks at Google. Almost everything we do involves words. Words are how we persuade, communicate, and connect.

They're how leaders lead, salespeople sell, and parents parent. But certain words are more impactful than others. They're better at changing minds, engaging audiences, and driving action. What are these magic words, and how can we take advantage of their power? Jonah joins Google to discuss his book, "Magic Words: What to Say to Get Your Way."

Jonah is a Wharton professor, internationally best-selling author and world-renowned expert on change, word of mouth, natural language processing and how products, services and ideas catch on. He has published over 70 articles in top-tier academic journals, teaches one of the world's most popular online courses and accounts of his work often appear in places like the New York Times, Wall Street Journal and Harvard Business Review.

Originally published in June 2023, here is Jonah Berger, what to say to get your way. So everyone here today actually has something very much in common, regardless what role you have within the organization, regardless of what your daily duties or jobs are. We all have something very much in common, which is that we all have something that we want to change.

For the folks that are externally facing, it might be a customer or a prospect's mind. For the folks that are internal, it might be a boss or a colleague. You might have something you want to catch on. You might have something, someone you want to influence. But the question is how we can do that more effectively. And today I'm going to focus on the power of language, the power of the words we use to do exactly that. Now, I don't have to tell you that we use language all the time.

We use language to write emails and put together PowerPoint presentations. We use language to give talks and send our thoughts via text. We talk to our friends, our colleagues, our bosses, our prospects, even our own private thoughts rely on language. But while we spend a lot of time thinking about the ideas we want to communicate, we spend a lot less time thinking about the specific words we want to use to communicate those ideas.

But unfortunately, that's a mistake because subtle shifts in the language we use can have a big effect on our own impact.

As I'll talk about today, for example, adding just two simple letters to a request makes people about a third more likely to say yes. When suggesting that we like something, saying we recommend that thing rather than we like that thing makes people about a third more likely to take our suggestions. And when we look at the language we use in email to communicate with our colleagues at work, the language we might use in applying for a loan, or a variety of other types of language, those

Those words provide deep insight into who we are and how likely we are to behave certain ways in the future. How similar your language is to your colleagues, for example, predicts whether you're more likely to get promoted or fired or whether you're likely to leave an organization to find a better job somewhere else. And the language people use in their loan applications provides subtle but important hints about how likely they are to default or repay their loan in the future. And so one question then is what are these magic words and how can we take advantage of their power?

And so that's a long, good introduction to a title slide, but it's nicely mentioned. I'm Professor Jonah Berger. I'm a marketing professor at the Wharton School at the University of Pennsylvania. And what I'm going to do today in the 30 or so minutes we have together is give you a brief tour of my recent New York Times and Wall Street Journal bestseller, Magic Words, What to Say to Get Your Way. I'm going to talk about the science of language and more importantly, how we can use that language. And

And I don't have to tell you that a lot has changed in the last few decades. More and more language data is available than ever before. From the emails we write, from the content people post online, even transcribing a conversation like this can give us easy access to language data.

And new tools have allowed us to parse this data in exciting and insightful new ways. Everything from old tools like dictionaries and topic modeling to new approaches like word embeddings and even more sophisticated tools have allowed us to unlock understandings about language that we never had before.

And so what I'm going to talk about today is some of what we can do with that language. In the book, I talk about two things. I talk about how we can use language to impact others and ourselves, and we can gain insight from language to understand our customers, clients, and others. I'm going to focus today on that first bucket, how we can tactically use language in our daily lives and our roles within the organization. But if you're interested, I'm happy during the Q&A to talk more about what language reflects,

and how we can use language data for insight, whether for customers or otherwise.

But in writing this book, I've studied an amazing set of different types of language. We've looked at tens of thousands of pieces of online content to understand what holds attention. We've looked at thousands of conversations, both between regular people and with companies, to see what increases customer satisfaction and what builds relationships between individuals. And in those studies and others, we found that there are six key types of magic words, six key types of language that we can use to increase our impact.

And in Magic Words, I put them in a framework, and that is the SPEAK framework. And that stands for the language of similarity and difference. The P is for the language of posing questions. The E is for the language of emotion. The A is for the language of agency and identity. The C is for the language of confidence. And the second C is the language of concreteness. And for those of you who are paying attention, you probably are going, "Speak doesn't sound-- it doesn't spell with two Cs."

you're exactly right. There should be a K. I was not clever enough to come up with speak with a K. I've been told, though, that K is the most difficult letter in Scrabble, so I don't feel so bad about that. But hopefully speak with two Cs will help you remember the framework. And so each of these is a type of language we can use in a variety of aspects of our personal and professional lives, not only to influence others, but to increase our

creativity and come up with better solutions to problems, to deepen social connection with our family members and peers, and have a host of other impacts across our daily lives. I won't have time today to talk about all six, unfortunately, but I'll dive into a little bit of the language of agency. I'll talk about the language of confidence because I think that's a fun one as well. I'll briefly touch on the language of posing questions, and then I'll wrap up.

But as was nicely mentioned at the beginning, I would love to take any and all questions. So please feel free to ask questions either along the way or at the end. If you don't get your questions in, you can find me at J1Burger, which is down on the corner of your screen on either LinkedIn or Twitter. And there are also a whole bunch of free resources about the book, about the framework, and how to apply these ideas at my website, which is in the other corner of your screen at jonaburger.com.

And so my goal today is very simple. I want us to better understand the language we use and how we can use it more effectively. And so let's dive in. And so I want to start with a common problem that many of us have. Often we want to get somebody else to do something, whether that person is a colleague at work, whether that person is a prospect, whether that person is a friend, whether that person is a child. We want to encourage them to do something in particular, but often it doesn't work.

Could we use language more effectively to do that? And so a few years ago, some scientists from Stanford University did a study to find out exactly that. They wondered if they could use language to increase persuasion. So they went to a local elementary school, preschool actually, and they asked four and five-year-old kids to help them clean up the classroom. It was a bit of a mess. There were blocks everywhere and there were crayons and books and various things that needed to be cleaned up.

And so for some of the kids, they asked like someone might usually ask someone. They said, hey, can you help clean up as a teacher or a friend might often ask. But for the other set of individuals, rather than asking for help, they asked for something slightly different. They said, hey, can you be a helper?

Now, as you can see on the screen, the difference between help and helper is quite small. It's only two letters. Yet those two letters had a very big impact. They led to about a 30% increase in the proportion of students that ended up helping to clean up. Now, that's an interesting study, but you might be sitting there going, well, those are kids. Those are classrooms. How do I know that this is work with adults with real consequential behaviors?

And so more recently, some researchers looked at the similar idea in a much more important context, and that is with voting. Obviously, voting is a really important behavior. It's a consequential behavior for our democracy, but it also requires a lot of effort. Not everybody wants to do it, and not everybody does. And so they wondered whether they could use language to increase the likelihood that people turned out to vote. So for some people, they sent them a message saying, hey, can you please go vote? And for other people, they said a slightly different thing. They said, hey, can you be a voter?

Now, here the difference is even smaller. It's just one letter. Yet this one letter led to about a 15% increase in people's likelihood of turning out to vote. And so what gives? Why was helper more effective than help? Why is motor more effective than vote? Well, it turns out the difference comes down to the difference between actions and identities.

We all know we should take certain actions, help, vote, do a variety of different things. But we're busy. We don't always have enough time. But what we care about more than holding desired actions and having desired actions is holding desired identities. We all want to see ourselves as competent, intelligent, smart, capable, and a variety of different things. And so we engage in actions that allow us to confirm those desired identities. If we want to see ourselves as athletic, for example, we need to go for a run once in a while.

But so by turning actions into identities, by framing actions into identities, we can make people more likely to take desired actions. Helping, sure, I know I should help. But being a helper, well, that's a desire of identity. I want to take that identity. And so to do that, I have to help. Same thing with voting. I know I should vote. I may not have the time, but being a voter, I definitely want to do that. And if voting is the way to show myself and others that I'm a voter, well, now I'm much more likely to do it.

But it's not just for positive things like voting or helping. It turns out the same thing holds for negative behaviors, but in the opposite direction.

Some research, for example, looked at negative identities and negative actions. Losing is bad, but being a loser, well, that would be even worse. Similarly, no one wants to cheat on a test, certainly a bad thing, but being labeled a cheater would be even worse. And research finds that when cheating would make you a cheater, how people are less likely to cheat on that test because they don't want to hold, take an action that leads them to see themselves as holding that undesired identity. It's almost like that old slogan,

getting people not to litter, they said, don't be a litterbug. Littering, I know I'm not supposed to litter. Well, hold on. If littering would make me a litterbug, now I'm a lot less likely to do it. And so by turning actions into identities, we can make people more or less likely to take those actions. And it turns out this idea goes even beyond that. It turns out the same thing can be used for describing people, including ourselves. Imagine I tell you about two friends of mine, one who runs and one who is a runner.

If you had to guess which of those two people would you say runs more? The person who runs or the person who is a runner?

We probably say, well, the person who is a runner, right? If they are a runner, it suggests something stable. It's not just transitory here today, gone tomorrow. It's a more consistent part of their identity. If someone drinks coffee, yeah, once in a while, they have a cup of coffee. If they're a coffee drinker, that suggests it's part of who they are. And so by describing people in terms of identities rather than actions, we can encourage them to take desired actions. Talking about ourselves, for example, as a runner, well, to hold that identity, we've got to run once in a while.

Similarly, if we describe a colleague, rather than describing them as hardworking, we suggest, yeah, they're pretty hardworking. We describe them as a hard worker, suggest it's a more consistent part of who they are. Trying to motivate someone to take an action, rather than telling them they should lead more often, tell them to be a leader. Framing these things as identities makes them seem more consistent and more long-lasting.

I particularly love the term creator, for example. Rather than telling, well, I create videos or I create content, which sounds a little bit like a side hustle, labeling that thing, I am a creator, I am an influencer. Now those things sound like full-time jobs, right? And because they sound like full-time jobs, they're a more stable aspect of who we are. And so these are just a couple of examples, but they're all examples of the same concept, which is the language of agency and identity.

Often we think about words as something we use to request information for others, but words do much more than that. Worms shape what it means to engage in a particular action and who's responsible or who's seen as blameworthy for doing something in particular.

Let me give you just one more example of this bucket. And this one comes from some research on problem solving. Often we're stuck in a difficult situation. We're trying to solve a problem. We're trying to come up with a creative solution. We don't know how to do it. We're sort of spinning our wheels. We're not really getting anywhere. We're often stuck. We're trying to figure out what to do. Well, often in those situations, we ask ourselves, what should I do? What should I do? How should I solve this problem?

But some researchers wondered if a slight shift in language could lead people to come up with better solutions. So they gave people a bunch of problems. And for some people, they asked them, what should you do in this situation? But for a second set of people, they used a subtle linguistic shift. Rather than ask them what they should do, they asked them what they could do in that situation.

Now, again, the difference between should and could is pretty small. It's only a few letters. Yet those few letters led to a big increase in how creative people were in solving those problems. People came up with better problems, better solutions to those problems, and rated as more creative the solutions when they were using coulds instead of shoulds. Because shoulds focus us on there's one particular answer. There's one right course of action. And if we could only get there, we will be successful. But coulds incurred us to widen our viewpoint a little

to realize there are multiple courses of action, multiple places we could go or things we could think about. And all of them may not be good ideas. All of them may not be the best possible approach. But at least by considering that wider set of options before coming up with a final solution, we're much more likely to reach a positive outcome in the end.

And so when stuck on a tough situation or trying to motivate others to come up with a better solution, don't just ask what we should do. Tell them and us to think about what we could do instead. By focusing on coulds rather than shoulds, we can come up with better solutions and be more creative as a result.

Okay, those are just a couple examples of the language of agency and identity. And in Magic Words, I talk about five or six other sort of sub buckets there of language we use. I talk about the power of you, for example, as a great word that grabs people's attention, increasing engagement on social media, increasing open rates of emails, but

making help content on like a customer service page, for example, less helpful because it suggests someone might be to blame for a particular action. So I'll let you check out some of that stuff in the book. I want to switch and spend a couple minutes on the language of confidence before wrapping up.

And so we all have someone in our lives who we think about as particularly charismatic, right? When they open their mouths, everybody listens. When they tell a story, everyone hangs on their every word. I know I certainly have people in my own life like that. I would love to be that person. I'm personally not like that. But one question I always wonder is why, right? What makes those people so charismatic? Why are they so effective in communicating?

And so to help us answer this question, I want to look at a communicator that many of us are familiar with, and that is none other than Donald Trump. Now, I don't want to get into politics today. You're entitled to whatever view you hold. If you like Donald Trump, that's just fine. If you hate Donald Trump, that's fine often. Fine also. I want to focus on his language in particular, because whether you like him or hate him, and particularly whether if you hate him, you can't deny that he's been amazing at selling his ideas.

Right. He's done a great job, particularly who hate his policies and ideas. You might wonder, how is it possible that he gets people to support them? But he has. And so one question is, what is he doing that makes him such an effective communicator, even if we don't like it or if you like it, too, that that's completely fine. So let's take a look at a speech he made when he originally started his presidential run many, many years ago.

So he said something like the following. I will build a great wall and nobody builds better walls than me. Believe me, and I'll build them very inexpensively. Our country's in serious trouble. We don't have victories anymore. We used to have victories, but we don't have them. When was the last time anybody saw us beating, let's say, China in a trade deal? I beat China all the time, all the time.

Now, when he made this speech, it was widely panned in the media. People said it was overly simplistic and full of bluster. Yet less than a year later, he was elected president of the United States. And so regardless of whether you like him or not, something he's doing is working. What is it?

Well, if you look, there's actually something that Trump does that he's not the only one that does. What he does is something that he has in common with great salespeople, with transformational leaders and entrepreneurs, folks like Steve Jobs or Elon Musk, and also gurus or others that are particularly charismatic. And that is that Trump speaks with a great deal of confidence.

He speaks with a great deal of what psychologists would describe as certainty. And what do I mean by that? Well, he uses a lot of what some people might describe as definites. He says things are definitely true. This is clearly the right answer. It's obviously a good course of action. Unambiguous, undeniable, guaranteed, precisely, essential. These are all definites. He speaks like things are absolutely, positively certain.

He doesn't just say something might happen. He says something that will happen. Take a look at that speech, for example. He doesn't say, I'll just build a wall, build a great wall. And nobody builds walls better than me. And he'll build them very inexpensively. He also doesn't say, hey, our country doesn't have as many victories as we used to. We don't have victories anymore. Absolutes. He speaks in absolutes. When was the last time you saw us beating China? I beat China all the time. He speaks like the answer is clear and obvious and absolutely true.

And it turns out that speaking with such certainty actually has an important impact on persuasion. Some research, for example, looks at the context of financial advisors. They ask people to choose between two advisors, one who spoke with more certainty and one who spoke with less certainty.

And it turned out that people preferred the advisor who spoke with greater certainty, even though that advisor was no less likely to be correct in their predictions. In fact, even though that advisor was more likely to be overconfident, when they got it wrong, they got it wrong by more. People still preferred the certain financial advisor because when people speak with so much certainty, whether a financial advisor, a guru, or Donald Trump, it's hard to believe that they couldn't be right because they seem so certain about what they're saying.

Now, contrast that with the way most of us speak most of the time. I don't know about you, but I tend to speak with a great deal of uncertainty. I often work with consulting clients and they ask me, hey, what do you think of this strategy? Or do you think this is the right idea?

I worked with the OnHub team, for example, at Google a number of years ago, and also with Project Aura and a variety of other groups, including Google Fiber. And I often say things like, well, this might work, or this seems like a good course of action, or I think this strategy is worth pursuing. What I'm doing there is what most people do most of the time. We hedge.

Hedges are simple linguistic turns of phrase that allow us to sort of balance things a little bit. If we're not exactly sure, if we don't know what we want to say, hedging is an easy way to sort of protect ourselves. But the problem is that by hedging, we are undermining our impact.

We often use it as sort of an easy phrase just because it's convenient. But we found in our research that the more we hedge, the more hedges we use, the less persuaded other people are to do what we want them to do. In part, because when we hedge, we seem less certain.

People are sitting there going, well, if you're not even sure it's going to work, why should we take this course of action? And even further, right, if you're not sure, maybe we should hire someone else who knows the right course of action. When we hedge, we undermine our impact. And so this has a few implications. First, we got to ditch the hedges.

There may be times where we want to express uncertainty, and I'll get to that in a couple minutes, but most of us just use hedges because they're convenient. Ditch the hedges. Unless we're trying to communicate uncertainty, don't hedge. Hedging undermines our impact even when we're not intending to do so.

Second, though, when we need to communicate uncertainty, we can name it. And let me show you what I mean by that. Imagine if I'm not sure a strategy will work. I'm not sure a particular direction is a good idea. I could say something like, I'm not sure this will work. Or I could say something like the following. Hey, I think this is a great idea. But for it to happen, these three things need to end up occurring. Notice the difference between these two things. In both of them, I'm expressing uncertainty.

But in one, I'm saying, I'm not sure it'll work. And in the other, I'm making it very clear that I think it's a great idea, but these things need to happen for it to work. That second way, I'm calling out the uncertainty. I'm identifying where the speed bumps or the roadblocks or the hurdles are. I'm not just saying, I don't know. I'm not just saying it's not clear. I'm saying, here's what I think, but we need to make sure these things occur. And so by naming or calling out, identifying the uncertainty, we can seem more confident and we can be more likely to persuade others.

We can also take ownership of the statements we're making. And I did that even on the last example I gave you. Contrast the difference between it seems versus it seems to me. In both cases, I'm expressing uncertainty, but in one of them, I'm owning that uncertainty. I'm taking ownership. I'm saying it seems to me like this is true. I'm not just saying this might work. I'm saying I think this might work.

I'm using a person, a first person singular pronoun like I or me by taking ownership of that uncertainty, by associating with ourselves. It makes us seem more confident and it makes other people more likely to listen to us as a result.

And also even beyond hedges, we want to avoid hesitating. We want to avoid using filler words, right? Often we use words like um, uh, or er. I do these things all the time. Some of us might use the word like. Why do we use these words? Well, we often use it to buy time to think. Particularly when we're speaking, particularly if someone asks us a question, we don't know exactly what we want to say. So we might say um or uh to buy ourselves time, to give ourselves time to think about what we want to say.

But the problem is while it may be a simple linguistic crutch that we use to buy ourselves time, it makes us seem less certain and makes other people less likely to listen to us. If you've ever listened to yourself speak, recorded yourself and listened to yourself, it's painful, but it's a great way to help identify those ums, uhs, which are called filler words.

Similarly, try to fill them in with pauses rather than fillers, right? Great speakers often pause. It's a nice way to draw the audience in and encourage them to pay attention to what we're saying. I have a bad tendency to speak quickly, particularly when I'm excited about something. But pausing is a great strategy that allows us to think rather than just filling it in with words and making us seem uncertain.

And further, if you're not sure whether using filler words, practice speaking, recording it, and spit out a transcription, right? It's a great way to easily identify the ums, uhs, and errs and help ourselves become better communicators.

And last but not least, the only thing I would say is I don't want you to walk away from this discussion going, well, wait a second. You're saying we should always communicate certainty because there are certainly times where uncertainty is more effective, right? If we don't know that something is true or we don't want to seem overconfident, communicating uncertainty can be a good thing to do. What I want us to understand is certainty is a tool, right? Linguistic certainty is a tool. There are times to be certain and there are times to be uncertain, just like there are times we need a hammer and there are times we need

screwdriver. But once we understand what certainty is and how to communicate it, we can deploy it in the right situations, right? If we're a leader, for example, or a salesperson, and we're getting up and making a big presentation or pitch, that's probably a time we want to seem pretty confident. On the other hand, if we're having a small team meeting, or we want to reveal that we're vulnerable or show that we're open to other people's opinions, well, then that's a time to hedge and use other markers of uncertainty.

Some very nice research, for example, shows that the more we use hedging and other types of uncertain language, the more open we seem to others' opinions and the more willing they are to listen to what we have to say because we're making it clear that we don't know the right answer.

And so there are times for uncertainty and there are times for certainty. Similarly, based on status within an organization, right? If we're a senior leader, sometimes we need to show a lot of certainty. On the other hand, if we're a junior member of a team and we've just joined, sometimes seeming overconfident can actually be detrimental. And so understanding how to communicate confidence and when to communicate it based on the context is a really important skill. Okay.

Okay, I give a couple more examples of the language of certainty and how to deploy it in that chapter on confidence. I want to give you one example of the language of posing questions, and then I'll wrap up and actually take some questions from you guys. And I find questions to be a really powerful rhetorical tool. Often we think of questions as a way to collect information, and they do that, but they do a lot more.

than just that, right? Questions shape how other people see us. They can act as spotlights, focusing attention on certain things rather than others. And they can be a great way to guide conversational flow and direct, whether it's conversation or attention, to certain things rather than others. So I think questions are a powerful tool in our communications arsenal that we don't always understand how to use. And so let me give you just one example of this. So

Often when we're stuck trying to solve a tough problem in a difficult situation, we often think about asking someone else for advice. We have a friend or a colleague who knows more than we do, and we want to ask them for advice, but we don't. And there are a few reasons why, right? First, we don't want to bother them. Second, we don't know whether they'll know the answer or not. But third, we're worried that they'll think less of us.

If I ask my boss or a colleague for advice, for example, they might think that I don't know what I'm talking about, that I'm not very good at my job. And so we're worried that there'll be a negative effect on how we're perceived based on asking for advice.

But it turns out that perception is actually a little bit misguided because some researchers from both Harvard and Wharton did a very nice study a few years ago where they had a variety of people have conversations and different types of interactions. And in some of those interactions, they had people ask their partner for advice. And then at the end of the interaction, they asked those partners what you thought of the person you were communicating with. And they found something quite interesting. It wasn't that asking for advice made us seem less intelligent or less competent.

And it wasn't that asking for advice had no effect. In fact, asking for advice had the exact opposite effect. It made people think we're more confident and more intelligent. And so you might be sitting there going, well, how could that be, right? How could asking for advice make us seem more confident and more intelligent? And the reason has everything to do with the fact that most people are pretty egocentric.

Think about ourselves. Most of us think we give great advice. Of course my advice is good advice. And so when someone comes along and asks us for our advice, we go, wow, they must be pretty sharp because of all the people they could ask for advice, they asked me.

And so as a result, they must be a pretty smart person. And so the moral here is really clear, right? Very simply, ask for advice, right? Whether you are a leader of a group or a team or whether you're just a member of that team, asking for advice not only allows you to collect valuable, useful and important information, but it can lead you to be perceived more positively as a result.

Now, there are a whole bunch of different lessons and takeaways and ideas in the chapter of posing questions, the how to ask better follow-up questions, when to ask certain types of questions versus others, how to use questions as deflection when people ask, put you in a difficult situation, to use that to shape the course of the conversation. Lots of useful stuff there, and I will let you check it out in the

in the questions chapter. But given I'm basically out of time, I'm going to do two slides, a wrap up, and then I will take your questions. And so first, today, very briefly, we've talked about six key ways to speak more effectively, six types of magic words.

In particular, we didn't get a chance to talk about the language of similarity and difference, but how similar our language is to others, both within an organization as well as outside of it, has an impact on our own effectiveness. Even research we've done looking at movie scripts, for example, shows that how similar chunks of movies are to one another impacts how engaging that content is. And so lots of take-homes in this chapter about how to write presentations, to put together content, and to do other things to hold attention.

We talked briefly about the language of posing questions and how to ask more effective questions. We didn't get to the language of emotion, but there's all sorts of interesting take-homes there about how to tell better stories, about how to pitch ourselves more effectively, and how to hold an audience's attention.

As I mentioned earlier, we did a big analysis of tens of thousands of pieces of online content, everything from news articles, front page news to things like lifestyle pieces, personal technology, blog posts, and measured how far down in those pieces of content people read to look at what leads to longer reads or how we can more effectively hold attention. So I'll let you check it out there. We talked a bit about the language of agency and identity.

how words shape how other people see us and how likely they are to take particular actions. We talk about turning actions into identities, turning shoulds into coulds. There are also some interesting take-homes there about words like you, I, and we, the power of talking to oneself to reduce anxiety and increase performance, and ways to motivate ourselves to achieve our goals by talking about having time, for example, and making time, or can't versus don't.

We talked briefly about the language of confidence and some tips there. And we didn't get to the language of concreteness, but there's some important insights in that bucket about how to show other people that we're listening.

We all know that listening is an important skill, but it's not just about doing the listening. It's about showing other people that we listened. And so concrete language can do actually exactly that. It can also help make our ideas more memorable and affect how likely we are as entrepreneurs to raise money or how good a leader we're seeing. And so I'll let you check it out in the concreteness chapters.

And so just to wrap up, two key next steps from this very short introduction to magic words. The first is we got to understand the power of language, right? We use words all the time. I've spent the last 25 minutes talking to you using words, both spoken as well as written. During this time, you may have taken notes using written words. You may have written emails or read other pieces of content that others produce. We use language all the time, but we don't necessarily know how to use it most effectively. By understanding the power of words, we can do that.

and then learn how to use them, right? By understanding these six type of words, we can be more effective. And as I mentioned, there's a resource tab on my site. Check that out. It has lots of useful guides for putting together more impactful content, being more persuasive, and so on. Because whether we know it or not, we are all speakers, right? We may not get up on a podium as a politician might and make a big speech, but we

get up in meetings and speak to teams and colleagues. We make phone calls and speak to peers and we talk to our friends. We're all writers, right? We may not write books or essays, but we write emails, we write presentations, we write text messages and all sorts of other content.

We may think that certain people are just born better speakers, better writers, better communicators, but it's actually not the case. It's a skill we can all learn. If we understand the six key types of magic words and how to use them, we can all increase our impact. So thank you so much. Hope you enjoy these ideas. Hope it's a good first step into your journey of applying language more effectively. And now I would love to take any and all questions. So thank you guys very much. Appreciate it.

Thank you so much, Jonah. I'd love to ask you a few questions about your creative process about actually writing these books. So to start off, like, do you write the books because the themes for the books emerge from the research you conduct or the research that you read? Or do you set up research in order to flush out ideas that you want to write a book about?

That's a great question. You know, I think it depends. Um, so what I, what I will say is that it's varied across books. So like contagious was, um, uh,

research that I had done in a class I was teaching and people were interested in the content, but it was only available in the class. And so I put it together as a book to sort of help people understand those ideas. The catalyst actually is some of my own research, but a lot of stuff that I was doing with consulting clients. And I just realized, hey, these things might make for a good book.

And magic words is kind of the stuff I've been working on for the last 10 or so years using natural language processing and automated text analysis, using everything from, you know, painfully basic tools like dictionaries. Some of you guys may be familiar with linguistic inquiry and word count or some of those basic tools. And more recently, more complex things like, you know, topic modeling and embeddings and so on.

I think there's a lot of powerful stuff there. I focus today on the really kind of usable things, you know, switching a word here and there. But I think some of the most insightful stuff is kind of parsing language for behavioral insight. And that's really where sort of large scale automated text analysis becomes useful. You talked a little bit about the computational tools for textual analysis. Yeah.

Are you writing those tools yourself? No, definitely not. So Jason mentions this very briefly, but he and I went to high school together many, many years ago at a school in the Washington, D.C. area called Montgomery Blair. And it was a math, science and computer science magnet school. So we took a lot of math, science, computer science classes.

I'm embarrassed to note, I think we use Pascal maybe, or maybe that was in middle school. And I used to have it on my resume and I got to college and my roommates mercilessly teased me for having Pascal on my resume. But I don't do, I'm not building new tools. I've spent time working with computer scientists. I actually visited Cornell Tech recently.

in New York City a number of years ago, working with some folks there. I'm really a behavioral scientist at heart. I'm really somebody who looks to understand human behavior, but I've just seen the power of these tools to allow us to parse textual data and now image and video data and other sorts of things for insights.

Just like the microscope revolutionized chemistry and biology and the telescope revolutionized astronomy, natural language and automated textual analysis tools are revolutionizing our understanding of human behavior. And so it's really opened up a lot of avenues for exciting research. Fantastic.

When I learn a new word, I often experience this confirmation bias. I see that word everywhere. After you were done writing magic words, did you find yourself noticing examples of the techniques you highlight in the book in your everyday life?

Some, yes. I actually start the story of the book with a story about our eldest son. I'll tell the story very briefly. A few years ago, our oldest son was born. Like many kids, he started using language. He had the word yo for yogurt. Brow bear was the way he described his brown bear. One of the most interesting words he used was the word peas.

And he meant please, but he didn't have his L's yet. So he had to say peas. But what was most interesting is the way he'd use this word, right? So he would say yo, if he wanted yogurt, and he'd look at you. And if you went to go get him yogurt, then that was it. But if you didn't get up and do immediately what he said, he'd try it again. He'd say yo, peas, right? And then he'd stare at you again, like, hey, you didn't do what I wanted. This word will get you to do what I wanted. And I found this super neat because it was the first time he had realized that words had power.

Right. That it wasn't just something you said, but the specific words you use could change the actions in others. And I think it's a really simple point, but a really important one. Right. The words we use when we communicate, whether through spoken or written language, have a big impact on how we're perceived and what others do as a result of what we say. And so by understanding those words, we can we can have a big effect on our own impact. Fantastic.

So you put the whole framework together as speak with two C's. How closely tied is speak to English?

Does it translate outside of English-speaking communities? Yeah, that's a great question. And so I was actually giving a presentation a few weeks ago to some journalists, and this very nice, polite German journalist at the end of the presentation sort of raised his hand. He said, I have a question. Are these things quantifiable?

culturally dependent because I was talking with someone in the US and I thought they had agreed to do an interview with me. But then I found out that they'd actually said no, but they had said it in such a sort of polite, careful way that I didn't even realize they had said no. I sort of thought they had said yes. And he's used to seeing German folks being very kind of direct and this is how it is. And Americans say, oh, thank you so much. I really appreciate it. And sort of hiding the no there. And

And so what I would say is this, you know, the words that express agency and identity are the same cross-culturally. The words that express confidence are obviously translated or similar cross-culturally. The question, though, is do we always want to communicate certainty, for example, may vary based on cultural context. America is obviously United States is obviously a very big culture of independence and sort of certainty and all those sort of things.

Whereas East Asian cultural context, maybe a little bit more about being a good member of a group and fitting in with that right group. And so they're, you know, being too confident and too certain, maybe perceived more negatively, at least if you don't have status within the group. And so I certainly think some of the types of language are the same, but how we deploy them, even within the U.S., right, depends on the context we're in and the sort of subcultures or cultural groups that we're a part of and we're communicating with.

Great. Well, we have some wonderful audience questions to add to the mix here. Tyler asks, "Any advice on how to navigate the balance between being confident and certain versus coming off as aggressive or a know-it-all?" Yeah. I think that is a key balance and I hope you heard as I was communicating confidence. I want us to help us understand what communicates confidence, but then we need to know how to deploy it.

And so I really like those sort of what I describe as personal hedges. I think it seems to me can be a nice way both to hedge, which indicates that there is some uncertainty, but also to own that uncertainty, right? Rather than saying, this is the best strategy, which can seem a little bit too much, right? Saying, I think this is the best strategy. This seems like the best strategy to me, right? It couches it a little bit, but it also shows that you're

You're taking ownership of it. You're respecting the fact that there may be diversity of opinions, right? I think this is the best strategy. That may mean you feel something different, but this is my opinion. And so I think situations like that or kind of owning that uncertainty by calling it out, right? I think this is the best strategy, but this is what we need to do to make it successful. Those are ways to make it clear that you've thought about it, right? You're not just sort of speaking off the top of your head because it's easy. You've thought about it, but

there is some uncertainty and this is the right way to navigate. Thanks. All right. Let's take another audience question. Another Jason asks, in a future in which our communications are polished by LLMs, what are, sorry, how much difference will an individual's own words or style matter?

Yeah, this has been the great part coming out with this book around this time, right? So it's funny. I was writing an article for a journal called American Psychologist with a colleague of mine probably about 12 months ago, maybe 18 months ago. And one of their viewers was like, oh, what about LLMs? And this was before all the chat GPT and other sorts of things. So we had a small section of the paper sort of talking about that. And now obviously everyone is wondering exactly this. I would say a few things. First,

These systems have done an amazing job at producing language, right? Of sort of creating language in ways we never thought possible, right? Realistic ways that feel like a person actually created it. And they're certainly going to overhaul certain industries, right? Industries that rely on sort of answering questions or creating content, they will be amazingly helpful.

What I think is less clear is how well they've been tuned for things like persuasion, how well they've been tuned for things like impact, at least so far. Eventually, they may be right. Eventually, someone may take enough of these and sort of have outcomes where they can say, look, this is more effective in changing something.

And so build that into or tune that into some of these systems. But I think at the moment, they're doing a better job of producing right answers or reasonable answers, right? Better than right answers is sort of reasonably seeming answers rather than producing the most persuasive answers. You know, they don't necessarily understand human behavior. Eventually, to tune correctly, they may do a good job of creating outcomes. But I don't think there's no place for understanding why people do what they

Great question, Jason. Thanks. All right. Next audience question from Kat. Have you explored the language brands use in their ads to connect with people or to get them to take action, et cetera? I have done a little bit of that. Yeah, I've done more work. So a couple of years ago, I analyzed, for example, all of brands' social media posts to understand what got more engagement and why, looking at linguistic features of those posts. As I sort of suggested earlier, I did another project looking at all their customer support pages.

So, you know, imagine you have a problem with your phone or your computer won't connect to something. You go to customer support. How is the way that page is written impact how helpful that content is perceived? I've done a little bit of work with advertising as well. And importantly, though, the language that works depends a little bit on the outcome.

So the language that grabs attention is not necessarily the language that people find the most helpful. Let's take the word you, for example. You does a great job of grabbing attention. When you're scrolling social media or deleting spam in your email, when something seems like it's self-relevant, it's focused on you, you're more likely to pay attention to.

And so, you know, compare five tips to save money with five tips to save you money, right? You're more likely to at least give that second one a second glance because it feels like it could be self-relevant. And so we find that content that uses second person pronouns like you, your, you'll, it's more likely to at least grab attention enough to increase engagement. That said, if you look at help content, for example, right? Customer support cage content, for example, you might think you is good there, but it actually has the exact opposite effect.

First, because they don't need to grab your attention. You've already gone looking for that page. They don't need to grab your attention. But second, you can feel like it's making someone feel blamed or like it's their responsibility. If you go to a customer support page, for example, says, if your phone is broken, you need to reboot the phone and then you need to do these six steps. Some might feel, well, why is it my responsibility to do all these steps, right? It's your piece of hardware that's broken, right? I paid you all this money for it. Why is it my responsibility?

And so kind of in thinking in those situations, you need to avoid accusatory use, right? You know, if you ask your colleague at work, oh, you know, did you finish that report? They could feel like, well, why is it my job to finish that report? And so you can suggest sort of responsibility or blame. It can also suggest agency. And so we need to be careful about how we use those types of words.

Wonderful. Thanks for that question, Cap. All right. Matilda asks, is there a fine line between persuasion and manipulation? You know, Matilda, I'm not sure you can still see the screen with all the titles of the books on it. But, you know, I've written a book called Invisible Influence. I've written a book subtitled How to Change Anyone's Mind. You know, what's tough about that question is

is if I was going to tell you that we're going to use language and influence the science of change to get people to eat healthier, to save the environment, to make better decisions, you would say they're fantastic.

On the other hand, if I told you we're going to use those tools to do bad things or, you know, hurt the environment, you'd say it's manipulation. It's negative. And, you know, the environment is shaping our decisions all the time. Right. If we go to a buffet line, for example, and somebody puts the vegetables in the first slot rather than the last lot of that buffet line, we're more likely to eat more vegetables.

Now, did that person manipulate us? They may have just been putting the vegetables where they happened to put them down, but that choice had an impact on our behavior. And so I think it's important to understand the behavioral science behind why we do what we do. We should certainly use that behavioral science ethically, but whether it's manipulation or not, I mean, we are being influenced all the time, even in unintentional ways. And so I think the more we understand how influence works, the more we can choose our influence.

and decide kind of when to be influenced and when not to be. Thanks for that question, Matilda. Time for one last question from the audience. Jenna asks, how did you choose what to include in Magic Words? What was the most difficult or surprising thing about writing your book?

I think what's tough for me is part of the stuff I find the most interesting here is not the most applicable. So I'll give you an example. We did a study of a quarter of a million songs over 60 years to measure misogyny in music, right? People say, oh, you know, hip hop lyrics are misogynist or this type of music is, is misogynist. Is it right? We measured the way songs talk about men and women over time to examine whether music is misogynist quick.

Quick hint, it's gotten less biased towards men over time, but it is still biased. And it's become less biased mainly because male artists have changed their lyrics. Female artists were not biased to begin with. I find work like that really fascinating. There's some research done by some great linguists and social psychologists at Stanford looking at the language of bias in a different context. They look at police language and find that police officers, even potentially unintentionally, speak differently to white versus African-American community members.

When traffic stops, it's important to understand kind of how language is used.

All of those are amazing insights. You know, we've looked at what makes movies successful or songs into hits based on language. I think all that stuff is amazing. It's not really things most of us can use in our daily lives. And so the first kind of two thirds of the book is more kind of tactical stuff you can use in meetings and writing emails to persuade, to be more creative and so on. The back half of the book is a little more kind of big picture, what language reflects about people, how we can predict whether someone's going to default on a loan based on

the language they use in their application, or how we can use language in some other exciting and innovative ways. And so I think the challenge is bridging those two things, because not only does language sort of impact others, it reflects things about the people that create it, the individuals or companies that create it. And so I think there's a lot of exciting things about language we can learn and understand.

Well, my thanks, Jonah, and to everyone in the audience today for tuning in and asking questions. Wherever you are in the world right now watching the stream, please take a moment to join me in thanking Dr. Berger for being our guest today on Talks at Google. His book, Magic Words, is in your favorite bookstores now and available digitally on Kindle and Audible. We'll see you again at the next Talks at Google.

Thanks for listening. You can watch this episode and tons of other great content at youtube.com slash talks at Google. Talk soon.