Jan 12

How and why we communicate with others

This is chapter 2 of my book Grouped. You you like and want to read the rest, you can do so here!





We talk to survive

The desire to communicate is hard-wired into all of us. It was an effective survival mechanism for our ancestors, who shared information about food supplies, dangerous animals, and weather patterns, and it continues to help us understand our world, including what behavior is appropriate and how to act in certain situations. People talk because sharing information makes life easier.

Our motivations for sharing online are the same as the motivations of our ancestors. We often update our status because we need information. Research has shown that the majority of tweets that mention brands are seeking information rather than expressing sentiment, and one in five tweets is about a product or service. (1)

We talk to form social bonds

Decades of research in social psychology has shown that people talk to form and grow social bonds. Conversations ensure that we understand one another. One key aspect of this is communal laughter. Research has shown that if people laugh together with strangers, they are as generous to them as they are to their friends. (2)

Talking to someone sends out strong social signals. It shows people that we consider them important enough to spend time together. This is also true online. People update their status to produce a feeling of connectedness, even when people are geographically distant. (3) Status updates often contain social gestures and people often respond by liking or commenting on the content, not because they actually like the content but because they want to send out a social signal to build the relationship. In many cases, the conversation that follows a status update is much more important than the status update itself. More than the act of sharing content, marketing campaigns need to support conversations.

Research has shown that social bonds are central to our happiness. The deeper the relationships someone has, the happier they will be. (4) Women talk to form social bonds more often than men. Many of their conversations are aimed at building and maintaining their social network. Men more often talk about themselves or things they claim to be knowledgeable about, often because they are trying to impress the people around them. (5)

We talk to help others

When researchers have studied why people share, they have consistently found that many do it to help others. This is an altruistic act with no expected reciprocity. For many, it is important to them to be perceived as helpful, and so they try to share content that they think other people will find valuable. (6) This is especially clear when we see people share information that may not reflect positively on themselves.

We talk to manage how others perceive us

While people talk to make their lives easier, to form social bonds, and to help others, most of our conversations are a form of reputation management. (7) Research has shown that most conversations are recounting personal experiences, or gossiping about who is doing what with whom. Only 5 percent is criticism or negative gossip. The vast majority of these conversations are positive, as we are driven to preserve a positive reputation. ( 8 )

Our identities are constantly shaped and refined by the conversations we have. Our values were passed on from conversations with our family, community, society, country, church, and through our profession, and are continually refined by the people we spend time with.

Quick tips

Build marketing campaigns that grow social bonds. For example, for Mother’s Day, the online florist 1-800-flowers.com used Facebook to have mothers vote on the products that they would prefer to receive as a gift. this generated stories in the News Feed, to be seen by their children.the motivation to grow social bonds led to four out of the five top-selling Mother’s Day products being the ones voted for on Facebook.

Build marketing campaigns that enable people to help each other. Sephora fans on Facebook organized to send each other unused cosmetics samples. One person starts a box of 30 samples, sends it to someone else who takes 15 samples out and adds 15 of their samples back, before sending it on to a third person who does the same, and so on.



Many of our conversations are about other people

One study on what people talk about found that about two thirds of conversations revolve around social issues. Another study found that social relationships and recounting personal experiences account for about 70 percent of conversations. Of the conversations about social relationships, about half are about people not present. The anthropologist Robin Dunbar described these conversations as “Who is doing what with whom, and whether it’s a good or bad thing, who is in and who is out, and why.” (5) Conversations about other people and their behavior help us understand what is socially acceptable in different situations by revealing how the people we’re talking to react to the behavior of the person not present.

Understanding how others have acted, as well as how the people we care about and trust react to those actions, shapes our behavior. It shapes what ideas we agree with, and how we may behave in the future. Supporting conversations about other people is critical for social products and for marketing campaigns based on social behavior.

We share feelings, not facts

Creative agencies the world over try to create content that people will spread. In order to do so, they need to understand what people share, and why. The vast majority of “viral” campaigns don’t spread at all, and this is often because the content is factual. Many research studies have shown that people don’t share facts, they share feelings. (9)

Jonah Berger and Katherine Milkman studied the most-emailed articles on the New York Times over more than a six-month period, totaling 7,500 items. They expected to find content that included factual information that might help others, such as diets or gadgets, but instead found that people shared the content that triggered the most arousing emotions. This included positive emotions such as awe, and negative emotions such as anger and anxiety. Emotions that were not arousing, for example sadness, did not trigger sharing of content. (10)

Content that is positive, informative, surprising, or interesting is shared more often than content that is not, and content that is prominently featured is shared more often than content that is not, but these factors are minor compared to how arousing the content is.

These findings have important implications for advertising. BMW ran a successful campaign called “The Hire,” which induced feelings of anxiety through elaborate car chases and generated millions of views. Content that is non-arousing, for example, content that makes people feel comfortable and relaxed, is unlikely to be shared. Public health information may spread more effectively if it induces feelings of anxiety rather than sadness. (11)

We talk about the things that surround us

Our everyday offline conversations tend to be about whatever comes to mind, independent of how interesting it is. And what usually comes to mind first is what is in our current environment (we’ll see later how this works for brands). If we’re talking to good friends, even our desire to appear interesting takes a backseat to environmental cues. Although we do craft our conversations in order to shape others’ perceptions of us,6 most day-to-day conversations with people we know well are about everyday things and are cued by our environment.

Conversely, our desire to appear a certain way to others is a bigger factor in what we talk about online than offline. Offline, many of our conversations are driven by a need to avoid awkward silences. While people most often talk about what is visible or cued by their environment offline, when online they don’t need to fill a conversation space so they can think more carefully about what might be interesting to others.

We talk about brands in passing

The research firm Keller Fay estimates that people talk about approximately 70 brands every week, an average of 10 a day. (12) We might imagine that people talk at length about the pros and cons of competing brands, but most of the time this is not so. Most references to brands in conversations happen in passing. People are talking about something loosely related to the brand, the brand comes up for a few sentences, and then disappears, as the conversation continues about the core topic. When people talk about brands, they are usually not motivated by the brand but by the instinct to converse with others and fill conversation spaces. We need to understand the incidental nature of brand conversations when planning marketing campaigns.

Research has shown that around Halloween, when there are more environmental cues about the color orange, products that are orange (Reese’s Pieces, orange soda) are more top of mind. (13) Other research found that products that are cued by the surrounding environment are talked about 22 percent of the time, versus 4 percent for products not cued by the environment. Products that are publicly visible are talked about 19 percent of the time, versus 2 percent for products that are not publicly visible. For example, in one research study, upcoming concerts were talked about much more often when there were CDs in the room. (14) We talk about eating much more often than technology or media, yet many assume that the latter are objectively more interesting.

This has profound implications for understanding how people talk about brands. Products that are visible and accessible will be talked about more. Products that are not naturally in people’s environment need to build associations with things that are in people’s environments. Yet, samples are not a substitute for the actual thing. Coupons and samples do not drive more conversations, but giving people the full product to try, so that it is consistently in the person’s environment, can lead to a 20 percent increase in conversations about that product. (14)

Interesting (arousing) products are talked about more initially, but once the novelty wears off, they are talked about less than things cued by people’s environments. Frequency of use also drives conversations, as products used frequently are easier to recall from memory and are therefore more top of mind. (15, 16, 17) People talk about big brands far more often than smaller brands. This is not surprising, as bigger brands are more accessible—more visible and easier to recall from memory.

Because we communicate much more frequently with the small number of people we are emotionally closest to, about half of conversations that mention brands are with a partner or family member. (12) Of these brand conversations, 71 percent are face to face, 17 percent are on the phone, and only 9 percent are online. (12) When it comes to spreading ideas, we need to target people’s closest ties.

Quick tips

Online posts that ask people to talk about others are likely to have high engagement rates. Many brands ask people to mention others in their responses, like this example from Jameson Irish Whiskey.

Polls are a great way to drive conversations about your business because the lightweight nature of interaction makes them more aligned with how brands bubble up and dissipate in natural conversations, like this example from Target.

Build campaigns around content that generates strong feelings, as it’s more likely to be shared. Marmite is a food brand in the uk that is either loved or hated by people.to generate sharing from the people who hate Marmite, they created a Facebook page called “the Marmite Hate party.”

if you’re trying to get people to talk about your brand, put it in their physical environment, as people will talk about things that surround them. Huggies had people upload their favorite photos of their babies to Facebook and then had the most popular photos printed on buses and in subway stations.



Most of our communication is with the people closest to us

We like to think that we talk to a wide and diverse set of people, but the reality is that we talk to the same, small group of people again and again. Research shows that people have consistent communication with between 7 and 15 people, but that most conversations are with our five strongest ties. We communicate with the same 5 to 10 people 80 percent of the time. (2) Keller Fay found that 27 percent of our conversations are with our spouse/partner, 25 percent are with a family member, and 10 percent are with a best friend. That’s 62 percent of our conversations with the people closest to us. Only 5 percent of our conversations are with acquaintances, and only 2 percent are with strangers. The remaining 31 percent is with the rest of the people in our social network. (12)

Research shows that people use social networks primarily to strengthen the bonds with their strong ties, and secondarily to build relationships with weak ties. When we looked at how many different people members communicated with directly on Facebook every week, including private messages, chats, wall posts, and likes and comments on status updates, we saw that the average was just 4 people. When we looked at how many different people they communicated with every month, it was only 6 people. This is despite the fact that these people are checking Facebook almost every day. (18) Other research has shown that the more people see each other in person or talk on the phone, the more they communicate online. (19)

We can map how frequently we communicate with others onto our social network structure:

We communicate more with the people toward the center of our social network, the people we are emotionally closest to.

Who is listening to us changes what we talk about

Who we talk to online has a large impact on what we talk about. Many people think carefully before posting status updates. Sometimes they have an explicit audience in mind for the post and need to consider whether it will be interesting or offending to the rest of the people they are connected to.

People are very conscious of being seen to be communicating information others will find interesting, funny, or useful. As they usually see only positive feedback, for example “likes” or comments on Facebook posts, it’s hard for them to know what other people find valuable. For many people the only way is to look at posts that receive no feedback, assume people didn’t find it interesting, and factor the characteristics of that post into future decisions about whether to post something. Sometimes people post updates broadly, as receiving serendipitous replies outweighs any risk of communicating uninteresting information to others.

We communicate differently to explicit groups of friends compared with larger groups of people.

When we talk in public, we’re very careful about what we say. For example, online public ratings tend to be disproportionately positive when they’re linked to our real identity. This is especially true when the other party involved can reciprocate. When people post anonymously, their ratings tend to be almost 20 percent lower than when they use their real names. When ratings are not visible to the party being rated, people give negative reviews more frequently. (20)

Quick tips

We need to build marketing campaigns around the people we’re closest to. When BMW launched the new Mini cooper in the us, they didn’t target people in the market for a car or people who fit their customer profile.they instead targeted existing Mini owners, as they knew that these people were the best way to influence their friends. (21)



People talk for a variety of reasons: Sharing information makes life easier, talking helps to grow social bonds with others, and choosing what we talk about allows us to manage how others perceive us.

We talk about other people, what’s around us, and things that generate strong feelings. Most conversations involve recounting personal experiences, or gossiping about who is doing what with whom.

We talk about brands in passing, often driven by what we see in our environment, and to fill a conversation space with someone else.

Most of our communication is with the people closest to us. We communicate with the same 5 to 10 people 80 percent of the time.



  1. See the 2009 research paper “Twitter power: Tweets as electronic word of mouth” by researchers at Pennsylvania State University and Twitter.
  2. See the 2011 research paper “Social laughter is correlated with an elevated pain threshold” by Robin Dunbar and others.
  3. See the 2010 research paper “Is it really about me? Message content in social awareness streams” by researchers at Rutgers University.
  4. For a great overview of research on happiness, see Derek Bok’s book The Politics of Happiness: What Government Can Learn from the New Research on Well-Being (Princeton University Press, 2010).
  5. See Robin Dunbar’s book How Many Friends Does One Person Need? (Faber and Faber, 2010).
  6. See the 2008 research paper “Word-of-mouth as self- enhancement” by Andrea Wojnicki and David Godes.
  7. For two examples, see the 1992 Social Psychology Newsletter article “The truth about gossip,” and the 1990 article “A social psychology of reputation,” both by Nick Emler.
  8. See Robin Dunbar’s book Grooming, Gossip, and the Evolution of Language (Harvard University Press, 1998).
  9. See the 2009 research paper “Emotion elicits the social sharing of emotion: Theory and empirical review” by Bernard Rimé.
  10. See the forthcoming 2012 research paper “What makes online content viral?” by Berger and Milkman.
  11. This example is from the 2011 research paper “Arousal increases social transmission of information” by Jonah Berger.
  12. The marketing consultancy Keller Fay have conducted many studies into how people converse. Explore their data at kellerfay.com/category/insights/.
  13. See the 2008 research paper “Dogs on the street, Pumas on your feet: How cues in the environment influence product evaluation and choice” by Jonah Berger and Gráinne Fitzsimons.
  14. See the 2011 research paper “What do people talk about? Drivers of immediate and ongoing word-of-mouth” by Jonah Berger and Eric Schwartz.
  15. See the 1977 social psychology research from Tory Higgins, William Rholes, and Carl Jones.
  16. See the 1982 research paper “Memory and attentional factors in consumer choice: Concepts and research methods” by John Lynch and Thomas Srull.
  17. See the 1990 research paper “Recall and consumer consideration sets: Influencing choice without altering brand evaluations” by Prakash Nedungadi.
  18. Statistics from internal analysis at Facebook.
  19. See the 2006 report “The strength of internet ties” by the Pew Research Center.
  20. See the 2010 research paper “I rate you. You rate me. Should we do so publicly?” by researchers at the University of Michigan, and the 2007 research paper “A familiar face(book): Profile elements as signals in an online social network” by researchers at Michigan State University.
  21. MINI’s innovative marketing strategy is described by Charlene Li and Josh Bernoff in their book Groundswell: Winning in a World Transformed by Social Technologies (Harvard Business Press, 2008).


Jan 12

A book is never done

Writing a book about social interaction is a strange experience because the minute the book hits the printing press, you start to uncover more and more new material that could have been included and you start to think about new things. I’m very lucky to be working in product development at Facebook, and am surrounded by some of the best minds in marketing and advertising. My daily conversations at work make me think about new areas, new combinations of ideas, new frameworks, and new possibilities for the future. For example, since the book was finished, I have started to read and think about the history of media technology, and how that might inform how people approach new media technologies today and in the future. I’ve also started to think a bit more about the relationship between homophily and influence. (This is largely because of some great conversations with my colleague Eytan Bakshy. Eytan is one of the brightest minds in this space and if you like this stuff, you should follow his work – he has some new papers coming out soon).

On my personal Twitter account, I’ve generally tried to keep the signal to noise ratio high (I post about my personal life on my Facebook account). I only tweet once every day or two despite reading tons of articles every day. This reading helps me form my thoughts, and as one of the most common pieces of feedback I’ve received about the book is that the references are really helpful, it made sense to create a new Twitter account for the book, where I will tweet the stuff I’m reading (with some quality control :). Think of it as a living extended bibliography for the book.

So, you can follow the book here @GroupedTheBook. I’ve dropped in a few articles I’ve recently to give you an idea of what I’ll post there. I’d love to hear thoughts if you read the articles I link to, and I hope to find new articles through the books’ followers. I hope you find the ongoing conversation helpful!

Jan 12

Shift 1 – Exponentially increasing information will dramatically change marketing

When I was growing up, we had a set of encyclopedias in our house. I often looked at these encyclopedias as a child, and thought that if I read all of them, I’d know almost everything (Of course, I never did read all of them). The access I had to information was bounded by what was printed, and available in local bookstores and libraries. The limit to the amount of information I could access was tangible. That world has irreversibly changed.

Thanks to the web, our access to information is increasing exponentially, and is not going to stop. A single query into a search engine produces millions of results. People are adding information to Wikipedia faster than we can read it: 1000 unique articles are added every day, on top of all the edits and additions to existing articles. We’re creating 2000 tweets per second, 48 hours of YouTube uploads per minute. In 2002, we recorded and replicated 23 exabytes of information. We now record and transfer that much information every 7 days.

This increase in the information we can access has two important side effects:

1. The capacity for information processing and memory is not increasing at anywhere close to the same rate. In fact, it’s barely increasing at all. Our capacity for processing information and remembering it, has formed over tens of thousands of years of evolution. In this sea of information, people often complain of “information overload” and although the amount of information we can now access is unprecedented, the feeling that we’re drowning in information is not new. In fact, it was first felt and described after the invention of the printing press. The word “deluge”, which means a sense of drowning, was a common metaphor for information in the 1600s.

2. This increase in information will have a profound impact on the world of marketing and advertising. Over the past 50 years, the predominant model in advertising has been to attract people’s attention – mostly by interruption, sometimes by surprising people with a juxtaposition of different aspects of society, creating new memes. Marketers and advertisers built systems that interrupted people from what they are currently focused on, to get them to focus on something else. We have long lived in a world where our TV shows are interrupted, our magazine articles are interrupted, our car journey is interrupted. If billboards were invented today, we wouldn’t know what they are because they would be banned by the road safety authority. The exponential increase in information we can access, combined with the fact that our brain has limited processing capability, means that a strategy based on interruption is now a race to the bottom. More advertisers, delivering more information, in an already overwhelming sea of information, competing for the same volume of attention. This ends badly for everybody.

While it’s important to understand the increase in information and the static nature of our brain’s processing capabilities, the critical thing to understand is how these relate to what is holding people’s attention. We’re watching less TV, and spending more time online. We’re spending less time online consuming content from broadcasters and businesses, and more time communicating with other people – most often with our friends. Advertisers are competing for people’s attention by going head to head with those peoples’ friends. It’s easy to see who wins here – we care a lot more about our friends and family than about what brands have to say, no matter how interesting or relevant the message. It’s annoying but acceptable to interrupt a TV episode of Friends, it’s not acceptable to interrupt a conversation amongst real-life friends.

In this eco-system, interruption is never a good experience. Imagine being at a party, talking to friends, and being interrupted by someone you didn’t know well, who immediately tried to change the topic of conversation to something that revolves around themselves. We have words to describe those kinds of people, and they’re not pretty.

Marketing and advertising is about to experience a paradigm shift. In 5 (maybe 10) years, we won’t see much advertising built on interruption because our information overloaded world will have rendered it ineffective and inefficient. All advertising will be context sensitive, personalized, and relevant. The distribution channels for this advertising are the technologies being invented now.

This post is the first in a series describing five huge shifts happening now.

Jan 12

Why the Olympics social media ban for volunteers is idiotic

Earlier on Twitter I posted that the social media ban on volunteers at the Olympics is idiotic. Let me explain why:

1. It’s a huge missed opportunity. Imagine if the organisers had decided to embrace social media from the volunteers. Imagine the moments that would be captured that couldn’t have ever been captured by the official TV crew. The best moments will be spontaneous and serendipitous. TV cameras won’t be there. Athletes winning, athletes losing. This highly emotional, and therefore engaging, content would have driven huge increases in interest because it would make the athletes more human, more like you and I, and would bring plenty of people in who will be on the fence about the Games. Many people in the UK are opposed to the Games. Many people globally have no interest in the Games. This increase in interest would create huge increases in viewership, and would build strong emotional relationships between ordinary people and the Olympics as a global event.

2. It’s not enforceable. People post under pseudonyms all the time. People can easily post anonymously to their friends who will then share the content. Loads and loads of content will leak out. Some people may be tracked down and fired, but most won’t be found. Remember, this is a group of 70,000 people who have no vested interest in TV rights, athlete rights, etc. They will be much more interested in sharing one of the highlights of their life with their family and friends.

3. It’s based on an understanding of a world which no longer exists. Any PR firm who believe that they can carefully control brand messages are deluded and are going out of business – slowly but surely. The role of PR has changed from command and control to engaging in conversation, and encouraging positive debate.

This is a really short sighted plan. I don’t buy that they are trying to protect the safety of athletes and VIPs. Are they also banning the Paparazzi? Are they worried that social media helps terrorists? In my opinion, this move is motivated by protecting the rights of those who paid extreme amounts of money to broadcast Olympic footage. It’s keeping all the major broadcasters happy. And maybe deep down they know the rule is not enforceable but have to toe the right line in public. But I worry that some people in the IOC with a lot of power have absolutely no idea how the world of media is changing.

I predict that we will see tons and tons of footage leaking out from the 70,000 volunteers, and that the best footage from the Games will come from regular folks, attendees and volunteers, and not from official TV crews. I also think that by the time the 2016 Olympics rolls around, this decision will be laughable, and the enforcers of this rule will look like dinosaurs.

Jan 12

Five major shifts

Lately, I’ve been talking to people about five major shifts that I see happening. Each is big enough to warrant a post of its own, so over the next few days I’ll write about each individually, and then write a post about what it means to think of them in combination. I’ll link to all from this post. In the meantime, here are the five major shifts:

1. The amount of information we can access is increasing exponentially.

2. The web is being rebuilt around people, rather than being built around content.

3. For the first time in humanity, social interaction, and influence, are measurable.

4. Technology is driving a large increase in understanding how we make decisions, and it’s not how we assumed.

5. Mobile technologies (phones, tablets, etc.) will change society in ways we can’t yet predict.