Nathan Bransford, Author


Tuesday, April 20, 2010

The Science of Buzz

Fads, crazes, hysterias, and other contagious social phenomena have long been a mysterious and occasionally hilarious part of human history. From meowing nuns to the Backstreet Boys, we humans have periodically been overtaken by curious mass obsessions for reasons that have never been entirely clear.

At least, until now. Science is catching up with buzz.

The Internet, as many a breathless commentator has told us, is greasing the wheels of the buzz machine like never before. The conversations and blog posts and general idea-spreading afforded by the Internet allows people around the world to instantly satiate society's pressing need for videos of children in a post-anesthesia haze and, or course, cute cats. Like a class of sneezing kindergartners, we are exposing and infecting each other with viral curiosities like never before, and this is allowing previously obscure media to come out of nowhere and catch on in a major way.

The mechanics of how buzz starts, keeps going, and eventually reaches saturation is not merely an object of curiosity for scientists. It's also big business.

Malcolm Gladwell, as is his wont, was at the cusp of addressing our Internet-era curiosity with how exactly this whole social viral process works in his book The Tipping Point. Gladwell identified what he saw as the mavens, the connectors, and salesmen who identified and spread social phenomena.

But while much of The Tipping Point was grounded in science, the process of how exactly an idea spreads from one person to the next was ultimately somewhat mysterious. For most of human history there was no real way of quantifying or tracking one person having a conversation with another person and saying, "OMG, there is this thing called a Slinky and it's seriously blowing my mind."

Now, however, we're having conversations on the Internet that everyone can see, record, and analyze.

During the aughts, as we were in the process of succumbing to a curious social contagion that led us to believe that stock markets reflected purely rational thought, we saw the rise of electronic futures markets to predict everything from presidential elections to whether Buzz Aldrin would win Dancing With the Stars. The theory being that if everyone is betting based on their individual knowledge and economic self-interest they will be able to predict the future better than an individual prognosticator/Nostradamus, an idea explored in James Surowiecki's The Wisdom of Crowds.

But wouldn't you know: it turns out that word of mouth can beat the market.

I recently had an opportunity to speak with Bernardo Huberman at HP's Social Computing Labs, who co-authored a fascinating study with Sitaram Asur that suggests the enormous potential in harnessing the power of the Internet to finally quantify word of mouth. Huberman and Asur tracked all the mentions on Twitter of upcoming movies leading up to their release, and they discovered that by measuring the number of times a movie was Tweeted they could predict a movie's opening gross more accurately than the Hollywood Stock Exchange.

Taking things one step further, they devised a method of tracking whether people were speaking positively and negatively about movies on Twitter after it came out. This improves the predictive power somewhat, but it turns out that Tweet-rate is still nearly as important after the movie comes out as before.

Their conclusion:

"While in this study we focused on the problem of predicting box office revenues of movies for the sake of having a clear metric of comparison with other methods, this method can be extended to a large panoply of topics, ranging from the future rating of products to agenda setting and election outcomes. At a deeper level, this work shows how social media expresses a collective wisdom which, when properly tapped, can yield an extremely powerful and accurate indicator of future outcomes."

It seems like a simple idea: of course people talking about a movie are more likely to go see it. But this is merely an initial step in quantifying buzz's power and how it spreads. Surely scientific analysis of stickiness, staying power, influence, advertising effectiveness, and Justin Bieber can't be far behind.

This could also potentially open up a world in which movies, books, music, and other media aren't simply dropped into the culture pond to see if they float, but rather they could be better marketed, and perhaps their eventual success more accurately predicted in advance. This could, at least theoretically, lead to some crucial reduction of uncertainty in businesses that have often been based on everyone's best guesses.

Not only is the Internet allowing us to exchange ideas faster than ever before, it's also allowing us to figure out the science of how word of mouth spreads. Buzz is going scientific.

Originally posted at the Huffington Post






39 comments:

Elizabeth Flora Ross said...

Great post, Nathan!

Anonymous said...

buzz kind of reminds of how in sports, even if it's just a pick-up game with friends, your team either has that mysterious, elusive momentum or not. and it makes all the difference in the game

jake

Bane of Anubis said...

Justin Bieber -- another reason to take over Canada

Anonymous said...

So now high-profile titles spend more of their big-budget marketing on blogs and internet marketing than on print and broadcast. Sounds about right.

Margaret Yang said...

If tweet rate=movie's success can be quantified, can it be manipulated? I suppose you could pay people to tweet about a movie, but others are going to tweet about it on their own. So, in a certain sense, word of mouth is still the best advertising that you can't buy.

Nathan Bransford said...

MY-

I think that's the next step - determining which kinds of Tweets are most effective at spreading and then utilizing those to generate actual buzz. Definitely fake buzz doesn't work, but surely there are ways of influencing the extent to which it catches on for real.

Christine Macdonald said...
This comment has been removed by the author.
Christine Macdonald said...

Great article. It will be insteresting to see where the scientific buzz takes us next. Hopefully more good than evil (marketing tools vs. cyber stalkers).

Chrstine

Kristin Laughtin said...

Fascinating.

Building off the idea that you can pay people to tweet, I imagine certain celebrities will be very effective. It could go either way with a corporate-based Twitter account, depending on how trustworthy the company is and whether their tweets are viewed as annoying or if people actually want to follow them.

Mira said...

You're so smart, Nathan.

I'm going to have to take awhile to think about what you said here, to absorb and understand it.

I did have a momentary chill though - when you said that obviously fake buzz won't work - are you sure about that(???)

Oh.

My.

Goodness.

Hmmmm.

Well, one thing that buzz does allow is the opportunity to fight fire with fire. That's comforting.

But we are entering interesting times. Thanks for having your finger on the pulse - admirable as always - and the implications here are fascinating.

Marilyn Peake said...

This is a fascinating topic, Nathan! With a Masters degree in Clinical Psychology and a Masters thesis rooted in Social Psychology, this topic completely fascinates me. The research experiment I conducted for my Masters thesis involved writing several versions of a fictional newspaper article about a person confessing to a crime while being tortured. All versions of the newspaper article were identical except for minor characteristics about the person or their situation. Subjects read one version of the article, then answered questions about whether or not the person was most likely guilty of the crime to which they confessed, how "moral" or "immoral" the person probably was, things like that. The results were shocking. People were automatically presumed to be innocent or guilty based on minor variations in the article that had nothing to do with any real facts about the crime. My Masters thesis advisor and I later presented a paper about my research at an annual meeting of the Eastern Psychological Association, and we spoke with members of Amnesty International who stopped by to discuss our results.

Right now, I believe the zeitgeist of our time is to mostly escape bad news. And media corporations with vested interests are more than happy to keep feeding the public what they want: entertainments that allow us to avoid more serious issues. As a result, the most popular TV shows and books are about gossip stuff. At some point, that will probably shift and then movies, TV shows and books that tackle reality will come back into vogue.

Reading piles of research reports on Social Psychology experiments has made it easier for me to step back from popular opinion and ask myself what are the real facts. Here are some fascinating experiments in which a majority of people will change their views even on objective facts (e.g. the length of a line or how far a small point of light appears to move) or moral decisions (whether or not to administer increasing strengths of electric shock to complaining victims) based on the opinion of the group or a strong authority figure:

Solomon Asch’s experiment on conformity regarding opinions about the length of a line

and

Milgram’s Experiment on Obedience to Authority

and

Muzafer Sherif’s "A Study of Some Social Factors in Perception", an experiment showing that a person will establish their own norm for how far a point of light appears to move, but will change their opinion based on peer influence.

mkcbunny said...

And soon you may be able to bet on those movie futures:

http://www.huffingtonpost.com/2010/04/16/regulators-approve-movie-_n_541085.html

Carl said...

Does the publishing industry currently do any kind of market research? Do they use focus groups, etc.

Anonymous said...

mkcbunny – Good Lord. Have you seen the movie, Capitalism: A Love Story by Michael Moore? He’s controversial, but the factual information in the movie is enough to make you cry. For example, lots of companies take out Dead Peasant life insurance policies on their workers through which the companies collect life insurance money even when the deceased family can’t afford to pay their deceased family member’s medical or funeral expenses. Is this mentioned prominently in the news? No. Do we know more than we ever need to know about celebrities from the news? Yes. But who wants to hear about depressing stuff like Dead Peasant policies? LaLaLaLaLaLa...Fingers in ears.

Marilyn Peake said...

mkcbunny,

I’m happy to see that the movie industry’s taking a stand against that, no matter what their reason might be for doing so.

Dara said...

Great post!

And before today, I had no idea who Justin Bieber was. I just knew he kept trending on Twitter and had no clue why :P

T. Anne said...

My daughter is in Justin B.'s demographic and it felt even to her that he came out of nowhere. He must have some amazing marketing machine behind him because his buzz preceded his success.

I know there are site's that will try to create a buzz for you, such as author buzz, and create buzz, but they can be expen$ive.

Great article Nathan.

Moira Young said...

Ah, I was wondering when someone would bring up all the "Beleibers" on Twitter.

It's an interesting phenomenon. If a picture is worth a thousand words, Twitter is a mosaic of photos where only 14% of the photo is visible - but that mosaic still forms quite an image of our collective thoughts.

Nonetheless, there's still netiquitte to be observed. Shameless self-overpromotion can have negative results because it turns people off. Harnessing the power of word-of-mouth is a delicate task.

Kelly said...

Thanks for posting on this, Nathan. Very interesting, even if I don't completely understand it (I do, however, know who Justin Bieber is).

Peter Dudley said...

Fake buzz can work initially because people can't tell immediately when it's fake. Look at IPOs in the dot-com boom. I knew a fund manager who said he would buy into the 3DO IPO (3DO was a game console manufacturer with a new idea) for the initial IPO bump even though the company and the product were not very impressive. This guy made a tidy profit for his funds while most of the people he sold the stock to after a few days ended up with worthless shares in a few years. My point is that IPOs during that time were created as much on buzz as on worthiness, and you couldn't always tell them apart until after the event.

Similar thing with premarketing books and movies.

But fake buzz will eventually undermine the credibility of the mechanism. If people can't trust the buzz, they'll start ignoring it, and it will lose its effectiveness. So, fake buzz will work for the lucky few that manage to make it happen now, but it will end up failing in the future.

Laurel said...

The buzz factor is lightening in a bottle, really. You can analyze the variables after the fact, but what makes it "buzz" instead of "formula" is whatever unquantifiable, unpredictable thing happened along at just the right time. Later you can see what it was. You just can't make it happen again since the freshness is gone.

The best you can do with manufactured hype is up the sales, which isn't bad. But a cultural phenomenon can't be duplicated in a lab because all the factors can never be accounted for.

Other Lisa said...

Can someone explain to me what a Justin Bieber is?

ryan field said...

There are a few celebs out there in the social networks who seem to know the trick to creating a buzz. It's interesting to watch and learn from them.

Renee Miller said...

Great post. Justin Beiber? I have an 11-year-old and a 5-year-old, both girls. I am Biebered out. I know his music and I don't want to. I memorized it under duress.

But he definitely had a machine behind him. Good for him I say. Just stop singing in my house.

Kairos Books said...

I think analyzing buzz is certainly worth doing, but I'm still not sold on how it can predict future buzz. Hindsight is 20/20, right? Explaining buzz is like explaining the reasons for historical events after they happen. But accurately predicting the future? Whole different ballgame.

Anonymous said...

Other Lisa - Here ya go ... Justin Bieber, teen pop sensation: Justin Bieber. Yawn...

Francis said...

A prepubescent boy couldn't have became such a sensation so fast 15 years ago. Internet really has opened new doors.

Mira said...

Okay, I recovered from my momentary fear about how this information might be utilized.

Bottomline, I believe that information should always be sought, regardless of the uses to which it might be put. After all, Darwinism led directly to the Holocaust, and yet would I want the entire theory of evolution to have been covered up?

Nope.

Information itself is innocent, and knowledge is powerful. That power can be used for good, for evil or, most frequently, for some combination, but that's part of our human journey, right?

That said, I think the implications here for research are truly awesome. Like I said in the forums, not only for studying 'buzz', but other social phenomena - like Marilyn pointed out, group think, obedience to authority and (another) mob mentality. Or peer pressure.

Maybe human interaction will be mapped in mathematical equations, that would be interesting.

Either way, the opportunity for us to understand ourselves better is priceless.

More priceless than a credit card. Ha, ha.

P.S. Wonderful article for the Huffington, Nathan. I'm admiring how smart you are, and if you think I'm sucking up by saying it twice, nope. This article just deserves having it said twice.

Gavin Brown said...

It's important to keep in mind that the tweets are a reflection of the buzz, not necessarily the cause.

Not only does fake buzz not help,
focusing huge resources on Twitter just because it's measurable may not be that effective either.

Ink said...

Otherlisa,

I think a Justin Bieber is when you fast-order a German lager. Right?

wendy said...

Incredibly well-written post. I thought I had some journalistic skills, but my efforts have been somewhat eclipsed. Sooo, the buzz is the best. I guess if anyone comes up with a commodity that people love then they're going to talk about it. Twitter or no. But I can see why its a good indicator of popular opinion.

I recently started sending out Tweets and am amazed that anyone would follow them. But I suspect these followers are mostly people selling things and are hopeful that I'll follow them and discover their product. They are usually motivational/guru types, perhaps because I pepper my tweets with
inspirational thoughts they see an interest.

Ishta Mercurio said...

Marilyn Peake, re: "As a result, the most popular TV shows and books are about gossip stuff. At some point, that will probably shift and then movies, TV shows and books that tackle reality will come back into vogue."

Interestingly, there has been an upsurge in recent years in movies based on real stories: Changeling, The Blind Side, Amelia, Flash of Genius, Rendition, Walk the Line... So maybe you're right. I certainly think that people are ready to start thinking about some of the hard questions in regards to the way the world works right now. And "David and Goliath" stories in which lowly citizens uncover the evils in government have always been popular.

Regarding the study of buzz leading to an effort to control it: don't they already do that? I think they call it "spin"...

As for me, I hope people will support my work because they like it and believe in it, and not because they've been convinced by marketing gurus that it's something they "just HAVE to have!" My favorite stories are the ones about indie films and small-press books that start with almost no marketing, but make it big because the people who see or read them just think they are that good and they tell everyone about them; consumer word-of-mouth (real buzz) is preferable to me over industry buzz and marketing (fake buzz, in my opinion).

Word verification: spent. Right on target. I swear, sometimes I think Blogger has an actual brain.

God and Ponytails said...

Something about me needs to believe the next new thing can make it just because it is awesome! Not because science has discovered the secret to buzz and we are all being manipulated into believing we love something.

I seriously loved the backstreet boys and now I love me some yogurt you get to top yourself. (Yogurt Planet: coming soon to a city near you)

csmith said...

Nathan,

If it is up your street, and you have time, have a read of "Pattern Recognition" by William Gibson. Deals with what buzz is, what creates buzz, and how buzz can be morally and ethically ambiguous in the extreme, all wrapped up in one of the strangest and most brilliant suspense/sci-fi stories I've ever read. I (heresy, I know) think it even surpasses Neuromancer.

All the best,

Chris

Nathan Bransford said...

csmith-

Ha. Yeah, cool hunting is kind of a literary agent's job. I read PATTERN RECOGNITION when it came out, though NEUROMANCER is still my favorite.

Kate Evangelista said...

I never really looked at Buzz that way before. The study was interesting. Actually, I was wondering what Justin Bieber had to do with your post, but now I know. Naughty.

J. T. Shea said...

Never for a single second, during the aughties or any other time, did I think the stock markets reflected purely rational thought! The essential delusion of many scientists, social scientists, economists, investors and other soft science 'experts' is that they can somehow predict what people are going to do BEFORE the people in question have themselves DECIDED what to do. I do believe there is a wisdom of crowds, but I don't believe it can be captured or controlled as easily as some 'experts' proclaim.
And no number of cute videos of Iggy the I-Pad Cat will ever persuade me to buy one. Well, maybe just one. I mean the cat of course! I'll NEVER buy an I-Pad! Well maybe just one...But no! I must not weaken!
Now where's Kindle Kat?

John said...

What's interesting is to run a Google search on a movie, etc. during opening night and watching the Twitter feed on Google.

The day Ricky Martin came out, there must've been a few hundred tweets rolling down the page in real time.

Sometimes you can see a large-scale buzz steamroller in action this way. Ah, the intarweb, my delectable series of pipes, how I love thee.

Steve said...

Okay,

Now that these predictive factors are sort of established, how long before twitter-spammers automate feeds of tweets optimized to game the predictive system?

"I'm certain that it happens all the time"

-Steve

Related Posts with Thumbnails