iCitizen 2008: Duncan Watts

40 %
60 %
Information about iCitizen 2008: Duncan Watts

Published on May 28, 2008

Author: iCitizen2008

Source: slideshare.net

Description

Keynote: Duncan Watts—Principal Research Scientist, Yahoo! Research

Influential or Insignificant?
Duncan rivets us with his empirical approach and application of network theory to sociology. So when he encourages brands to extend the conversation beyond the elite few and interact with networks on a vaster, exponential level, we can't help but listen.

DUNCAN WATTS Word of Mouth for the Real World

PEOPLE INFLUENCE EACH OTHER IN MANY WAYS MARKETERS HAVE ALWAYS EXPLOITED THESE EFFECTS Branding Endorsements Celebrities “Ordinary people” Product Placements Social Proof Being talked about in the media SOCIAL INFLUENCE IN MARKETING

PEOPLE INFLUENCE EACH OTHER IN MANY WAYS

MARKETERS HAVE ALWAYS EXPLOITED THESE EFFECTS

Branding

Endorsements

Celebrities

“Ordinary people”

Product Placements

Social Proof

Being talked about in the media

Research in 1950’s emphasized importance of personal influence Trusted ties more important than media influence in determining individual opinions Also found that not all people are equally influential A minority of “opinion leaders” or “influentials” are responsible for influencing everyone else Influentials, in turn, influenced by the media Together, these findings led to the “two-step flow” of influence Call this the “influentials hypothesis” BUT OF WORD OF MOUTH IS SPECIAL

Research in 1950’s emphasized importance of personal influence

Trusted ties more important than media influence in determining individual opinions

Also found that not all people are equally influential

A minority of “opinion leaders” or “influentials” are responsible for influencing everyone else

Influentials, in turn, influenced by the media

Together, these findings led to the “two-step flow” of influence

Call this the “influentials hypothesis”

THE “INFLUENTIALS HYPOTHESIS” “ One in ten Americans tells the other nine how to vote, where to eat, and what to buy.” (Keller and Berry, 2003)

IT’S A GREAT STORY… = “ Social epidemics ... are also driven by the efforts of a handful of exceptional people” Gladwell (2000) Free!!

East Village Hipsters start wearing Hush Puppies, and they become popular again Some unknown designer becomes famous when an actress wears his dress to Oscars Jeff Jarvis complains about Dell on his blog, and suddenly there’s an uproar A single patient in a Hong Kong Hospital jump starts the SARS epidemic AND IT SEEMS TO EXPLAIN A LOT

East Village Hipsters start wearing Hush Puppies, and they become popular again

Some unknown designer becomes famous when an actress wears his dress to Oscars

Jeff Jarvis complains about Dell on his blog, and suddenly there’s an uproar

A single patient in a Hong Kong Hospital jump starts the SARS epidemic

“ Influencers have become the ‘holy grail’ for today’s marketers.” —Rand (2004) But grails are hard to find…

The Influentials Hypothesis is simple and appealing A few special people (“Connectors,” “Mavens”) generate a huge impact It is also seemingly prescriptive “ Find the Influentials and Influence them” It should work everywhere It seems to explain so many things So why is it so difficult to implement? WHAT’S THE PROBLEM?

The Influentials Hypothesis is simple and appealing

A few special people (“Connectors,” “Mavens”) generate a huge impact

It is also seemingly prescriptive

“ Find the Influentials and Influence them”

It should work everywhere

It seems to explain so many things

So why is it so difficult to implement?

... but they operate in very different ways. Lots of kinds of people could be (and have been) called “influentials”...

If who is influential depends on circumstances, can anyone be an “influential”? Even ordinary people can be influentials in many different ways. Alpha Moms Connectors Brand Enthusiasts Mavens Trendsetters

Stories are important for making sense of things “ Hipsters wore hush puppies and then other people did. Therefore hipsters made hush puppies popular” What seemed mysterious is made to seem sensible, inevitable But, Stories only ever “explain” events in hindsight We only try to explain “interesting” outcomes Hipsters wear clothes every day Only need to account for one sequence of events Hush Puppies might have caught on anyway An explanatory theory also requires accounting for Everything that might have happened, but didn’t Everything that might have led to what did happen To make use of social influence, we need theories, not stories STORIES ARE NOT THEORIES

Stories are important for making sense of things

“ Hipsters wore hush puppies and then other people did. Therefore hipsters made hush puppies popular”

What seemed mysterious is made to seem sensible, inevitable

But, Stories only ever “explain” events in hindsight

We only try to explain “interesting” outcomes

Hipsters wear clothes every day

Only need to account for one sequence of events

Hush Puppies might have caught on anyway

An explanatory theory also requires accounting for

Everything that might have happened, but didn’t

Everything that might have led to what did happen

To make use of social influence, we need theories, not stories

Some people are more influential than others But no-one influences everyone Influence is Bi-directional Distributed Multi-step A THEORY OF SOCIAL INFLUENCE

Some people are more influential than others

But no-one influences everyone

Influence is

Bi-directional

Distributed

Multi-step

Computer simulations of influence networks Whether or not influence can spread widely depends mostly on the network structure If network permits spread, anyone can start something; and if not, no-one can Influentials at best modestly better starting points than average people Large cascades driven by “easily influenced individuals influencing other easily influenced individuals” Not “influentials influencing followers” SOME RECENT RESEARCH (Watts and Dodds, JCR, 2007)

Computer simulations of influence networks

Whether or not influence can spread widely depends mostly on the network structure

If network permits spread, anyone can start something; and if not, no-one can

Influentials at best modestly better starting points than average people

Large cascades driven by “easily influenced individuals influencing other easily influenced individuals”

Not “influentials influencing followers”

No-one would claim that large forest fires are started by “special” sparks Yet for social phenomena, we want to believe “special” outcomes are caused by special people A network view of influence suggests that individuals who later seem influential may simply be accidents of circumstances Obvious in hindsight, but not in advance FOREST FIRES AND ACCIDENTAL INFLUENTIALS

No-one would claim that large forest fires are started by “special” sparks

Yet for social phenomena, we want to believe “special” outcomes are caused by special people

A network view of influence suggests that individuals who later seem influential may simply be accidents of circumstances

Obvious in hindsight, but not in advance

Experimental study involving 15,000 subjects Designed to test how Individuals are influenced by what others think Subjects were shown a grid with MP3s from unknown bands They listened to, rated and downloaded favorites Behavior was tracked in several different “worlds” to measure social influence MORE UNCERTAINTY: THE MUSIC LAB EXPERIMENT

Experimental study involving 15,000 subjects

Designed to test how Individuals are influenced by what others think

Subjects were shown a grid with MP3s from unknown bands

They listened to, rated and downloaded favorites

Behavior was tracked in several different “worlds” to measure social influence

THE MUSIC LAB SETUP

Individuals clearly influenced by the votes of their peers Popular songs became more popular, and unpopular songs less popular, than in the independent condition Social influence increases inequality But also became harder to predict which particular songs would become popular Social influence increases unpredictability “Best” songs never do terribly and the “worst” never excel; but anything else is possible EXPERIMENT RESULTS (Salganik, Dodds, and Watts, 2006)

Individuals clearly influenced by the votes of their peers

Popular songs became more popular, and unpopular songs less popular, than in the independent condition

Social influence increases inequality

But also became harder to predict which particular songs would become popular

Social influence increases unpredictability

“Best” songs never do terribly and the “worst” never excel; but anything else is possible

When people influence each other, outcomes are inherently unpredictable . What we learn from the past is of little use in predicting or planning the future. No easy solution to this one. Be sceptical of holy grails and free lunches. But can identify some principles for operating that don’t depend on locating “special” people or having brilliant instincts. The Network Challenge: Radical Uncertainty

Aim for Easily Influenced masses over Influential Minority “ Bored at work network” is millions of workers who share media, blog, and IM all day Can make anything “go viral” if they like it But hard to predict what they’ll like Although quick, fun, and easy to share tends to do better So generate lots of options and /or variations Measure performance in real time Redirect energy/attention to successful ones Whenever possible, experiment PRINCIPLE #1 MEASURE AND REACT

Aim for Easily Influenced masses over Influential Minority

“ Bored at work network” is millions of workers who share media, blog, and IM all day

Can make anything “go viral” if they like it

But hard to predict what they’ll like

Although quick, fun, and easy to share tends to do better

So generate lots of options and /or variations

Measure performance in real time

Redirect energy/attention to successful ones

Whenever possible, experiment

Most things will not “go viral” Especially if your message is utilitarian, serious, and/or complicated Relying on small number of influentials simply aggravates the unpredictability Instead, target large number of ordinary individuals, and help them share message Target “big seed” of 10,000 people They recruit 5,000 extra people, Those people recruit 2,500, etc…, Eventually dies out, but get 10,000 extra in process PRINCIPLE #2 DON’T COUNT ON “TIPPING”

Most things will not “go viral”

Especially if your message is utilitarian, serious, and/or complicated

Relying on small number of influentials simply aggravates the unpredictability

Instead, target large number of ordinary individuals, and help them share message

Target “big seed” of 10,000 people

They recruit 5,000 extra people,

Those people recruit 2,500, etc…,

Eventually dies out, but get 10,000 extra in process

Using ForwardTrack Software developed at Eyebeam Tide Cold Water seeded with over 900K and got 40K extra Oxygen Media turned seed of 7,064 people Into 30,608 No “tipping” but clear, measurable ROI “ BIG SEED MARKETING”

Using ForwardTrack Software developed at Eyebeam

Tide Cold Water seeded with over 900K and got 40K extra

Oxygen Media turned seed of 7,064 people Into 30,608

No “tipping” but clear, measurable ROI

“ Mullet Strategy” Business up front, party in the back Let a thousand flowers bloom,then pick the best Huffington Post, Digg, etc. “ DIY Influentials” Pick the individuals who are already broadcasting your message, and make them influential BuzzFeed.com is platform for detecting and directing buzz PRINCIPLE #3 MANAGE, DON’T DICTATE

“ Mullet Strategy”

Business up front, party in the back

Let a thousand flowers bloom,then pick the best

Huffington Post, Digg, etc.

“ DIY Influentials”

Pick the individuals who are already broadcasting your message, and make them influential

BuzzFeed.com is platform for detecting and directing buzz

BuzzFeed embodies network thinking Locate Buzz Filter the Entire Web Trend Detector Crawler Search Manage Buzz Publish and Seed Measure and React Respond to Real Data Search Engine Tracking “ % Viral” Tracking Widget Click Stats The Web App: Publish Aggregated Buzz Widgets on Network BuzzFeed.com BuzzFeed Ads

Locate Buzz

Filter the Entire Web

Trend Detector

Crawler

Search

Measure and React

Respond to Real Data

Search Engine Tracking

“ % Viral” Tracking

Widget Click Stats

The Web App:

Publish Aggregated Buzz

Widgets on Network

BuzzFeed.com

BuzzFeed Ads

Bad news is that complexity of influence networks means we can’t predict either what will succeed, or who will make it succeed Good news is that we don’t need to Build Portfolios - measure and experiment Contagious Media - focus on the easily influenced Big Seed Marketing - viral marketing without tipping Mullet Strategy - try everything and promote what works DIY Influentials - promote those who promote you Main point is that network thinking replaces instinct The more we can measure, the more this will be true SUMMARY

Bad news is that complexity of influence networks means we can’t predict either what will succeed, or who will make it succeed

Good news is that we don’t need to

Build Portfolios - measure and experiment

Contagious Media - focus on the easily influenced

Big Seed Marketing - viral marketing without tipping

Mullet Strategy - try everything and promote what works

DIY Influentials - promote those who promote you

Main point is that network thinking replaces instinct

The more we can measure, the more this will be true

REFERENCES D. J. Watts and P. S. Dodds. Networks, influence, and public opinion formation. Journal of Consumer Research, 34(4), 441-458 (2007). D. J. Watts and J. Peretti. Viral marketing in the real world. Harvard Business Review (May, 2007) D. J. Watts. Is Justin Timberlake a product of cumulative advantage? The new theory of the hit record. New York Times Magazine (15 April, 2007) D. J. Watts. The Accidental Influentials. Harvard Business Review , p. 22-23 (February, 2007) D. J. Watts and S. Hasker. Marketing in an unpredictable world. Harvard Business Review , p. 25-30 (September, 2006)

D. J. Watts and P. S. Dodds. Networks, influence, and public opinion formation. Journal of Consumer Research, 34(4), 441-458 (2007).

D. J. Watts and J. Peretti. Viral marketing in the real world. Harvard Business Review (May, 2007)

D. J. Watts. Is Justin Timberlake a product of cumulative advantage? The new theory of the hit record. New York Times Magazine (15 April, 2007)

D. J. Watts. The Accidental Influentials. Harvard Business Review , p. 22-23 (February, 2007)

D. J. Watts and S. Hasker. Marketing in an unpredictable world. Harvard Business Review , p. 25-30 (September, 2006)

Add a comment

Related pages

Duncan J. Watts - Wikipedia, the free encyclopedia

Duncan J. Watts Duncan Watts; Watts ... Watts presenting at iCitizen 2008. Born: Duncan James Watts February 20, ... Duncan Watts page at Microsoft.
Read more

Duncan J. Watts - WOW.com

You have no Favorite Channels. To follow a channel click the ... If you wish to view your Favorite Channels from anywhere on the site, click on the My ...
Read more

Resource Interactive - HubSlide

iCitizen 2008: Duncan Watts Keynote: Duncan Watts—Principal Research Scientist, Yahoo!... 2 months ago. Business. iCitizen 2008: Steve Knox
Read more

Duncan J. Watts | QuickiWiki

Duncan Watts; Watts presenting at iCitizen 2008. Born: Duncan James Watts February 20, 1971 (age 45) [1] Guelph, Ontario [1] Residence:
Read more

Influence research: Duncan Watts and the debate on whether ...

iCitizen 2008: Duncan Watts. View more presentations from Resource Interactive. ... For the most current insights and trends in the living networks, ...
Read more

About: Duncan J. Watts - DBpedia

Duncan J. Watts, род. 1971) — американский ученый, ... Watts presenting at iCitizen 2008 (en) dbp:dateOfBirth: 1971 (xsd:integer)
Read more

DUNCAN WATTS - Escola de Redes

The Author Duncan Watts is a principal research scientist at Yahoo! Research, where he directs the Human Social Dynamics group. Prior to joining Yahoo!, he ...
Read more

Lectures & Appearances | Everything is Obvious

2008. iCitizen 2008: Duncan Watts. Slideshare. Big Thinkers India Series – Duncan Watts. Yahoo Research. 2004 “Six Degrees: Seminar with Duncan Watts. ...
Read more

Icitizen | LinkedIn

View 145 Icitizen posts, presentations, experts, and more. Get the professional knowledge you need on LinkedIn.
Read more