Word-of-Mouth Research: Principles
and Applications
Wordofmouth(WOM)isanimportantcomponentofacomplexanddynamic
marketplaceenvironment,andassuch,WOMresearchisbestundertakenaspartof
aholisticresearchprogram.FiveprinciplesdescribingtheoperationofWOMare
discussed,supportedbydata,andexamplesdrawnfromrecentresearchstudies.
Complexitysciencemodelingisintroducedasaneffectivemethodforsimulatingthe
real-worldoperationofWOMinagivenmarketcategoryandidentifyingwaysinwhich
marketerscaninfluenceittotheiradvantage.KeybusinessissueswhereWOM
researchcaninformdecisionmakingarelisted.
INTRODUCTION
Abundant research demonstrates that word of
mouth(WOM)isoneofthemostinfluentialchan-
nels of communication in the marketplace. The
reasonsfor WOM’s power are e evident: word of
mouth is seen as more credible than marketer-
initiatedcommunicationsbecause it is perceived
as having g passed through the unbiased filter of
“people like me.”Atatime ofdecliningtrust in
institutions, research shows that its influence is
growingstronger.
Inarecent nationalsurvey (HarrisInteractive,
2006a),U.S.consumers were e asked which infor-
mation sources they find useful l when deciding
whichproductstobuy infour commonproduct
categories. WOM and “recommendations from
friends/family/people at work/school” were e by
farthemostinfluentialsourcesforfastfood,cold
medicine,andbreakfastcereal.Forpersonalcom-
puters,ahighlytechnicalcategory,wesawastrong
reliance onexpert adviceintheform ofproduct
reviewsand websites,followed by WOMas the
nextmost useful.
While WOM has always played animportant
role intheformationofconsumeropinions,over
the past decade it has become an even more
powerfulforce,duetoatechnology-drivenexplo-
sioninthenumberandtypesofinformalcommu-
nicationchannels.Email,theinternet,cellphones,
PDAs,textmessaging,instantmessaging,andblogs
havemadesharinginformationandopinionseas-
ierthaneverbefore.Table1,basedontheAnnual
RQ
SM
(Reputation Quotient) study from Harris
Interactive (HarrisInteractive,2006b),showsthe
penetrationofseveralnewmediachannels.
It does not t take asophisticated d research ap-
proachtoconfirmthatWOMplaysaroleinagiven
category.ButtounderstandhowWOMoperatesand
why—soyoucanleverageittoyouradvantage—
requiresdiggingdeeper.WewillarguethatWOM
isa complex phenomenonthat must t be under-
stoodnotinisolation,but inthe contextofady-
namicmarketplace.Assuch,WOMresearchisrarely
astand-aloneeffort,butratherpartofaprogramof
researchtoaddressabroaderbusinessproblem.
ResearchandanalysesofWOMisstillanemerg-
ingfield.Overthepastfewyears,socialscientists
andmarketingpractitionershavemadeimportant
strides in describing the e components and struc-
tureofWOMinteractions.
Our focusinconsulting withclientsonWOM
for a number of years has been to o provide our
clientswithinsightsintothecomponentsofWOM
thataremostimportanttobusinessproblemsand
intomeasuringthosecomponentseffectively.Based
onthisexperience,wehaveformulatedapointof
viewandanapproachtomeasuring andanalyz-
ingWOMactivities.
DEET.ALLSOP
HarrisInteractive
dallsop@harris
interactive.com
BRYCER.BASSETT
HarrisInteractive
bbassett@harris
interactive.com
JAMESA.HOSKINS
HarrisInteractive
jhoskins@harris
interactive.com
398
JOURNALOFADVERTISINGRESEARCH
December 2007
DOI: 10.2501/S0021849907070419
Pdf link - insert, remove PDF links in C#.net, ASP.NET, MVC, Ajax, WinForms, WPF
Free C# example code is offered for users to edit PDF document hyperlink (url), like inserting and deleting
clickable links in pdf; add links in pdf
Pdf link - VB.NET PDF url edit library: insert, remove PDF links in vb.net, ASP.NET, MVC, Ajax, WinForms, WPF
Help to Insert a Hyperlink to Specified PDF Document Page
add links to pdf in acrobat; add url pdf
Thisarticlewill:
• outlinesomegeneralprinciplesbywhich
WOMoperates,
• sharesomedatafromongoingresearch
tobringsomeoftheseprinciplestolife,
• present a framework for understand-
ingthemajorcomponentsofWOM,
• describethe use ofcomplexityscience
modelingtechniquestomeasuretherel-
ative impact of WOM andto o identify
actions that will maximize positive
WOM,and
• identifybusinessproblemsthat t canbe
solved through WOM research and
analysis.
THEORYANDFOUNDATION
Theconsumer marketplace inwhichany
enterpriseoperatesisacomplex,dynamic
system.Word ofmouthplaysanimpor-
tantroleinthissystem,butitisonlyone
of many y things going g on, including g fac-
torsyou control (suchasmarketing and
promotion)and others youcannot t con-
trol(like the economy andcompetitors).
Starting with thispanoramicview helps
us set realistic expectations about what
canandcannot beaccomplishedthrough
WOMmarketing.
The successofthe enterprise depends
on building bonds (offamiliarity,favor-
ability,loyalty,etc.)withitskeystakehold-
ers,firstandforemostitscustomers.Those
individualstakeholders,inturn,takepart
in multiple social networks, where they
influence each other (throughWOM) in
theformationofattitudesandbehaviors
thatcaneitherstrengthenorweakenthese
bonds. It is critical for r the enterprise to
understandthe socialnetworksto which
itsstakeholdersbelongandhowtheyop-
erate, so it can influence the spread of
positiveWOMandminimizethedamage
ofnegativeWOM.
AswethinkaboutWOM,weareguided
bythefollowingprinciples.Keepingthem
firmlyinmindcanhelpbusinessesmake
better decisionssurrounding WOMand
whattheyshoulddoaboutit,ratherthan
blindlyjumpingonthe“buzz”bandwagon.
PRINCIPLE#1
Notallsocialnetworksareequal,andnot
all individuals in n a given socialnetwork
haveequalinfluence.
We have all seen headlines suggesting
that1in10Americansinfluencestheopin-
ionsoftherestofthepopulation.
Inhispopularbook The TippingPoint,
MalcolmGladwellwroteaboutthreeper-
sonalitytypes(mavens,connectors,andsales-
men) who play a key role in causing
messagestospread(Gladwell,2000).
While there do seem to be some of
these “special” individuals, their exis-
tence cannot fully explainthepervasive-
nessorthemechanicsofWOM.AsDave
Balterwritesinhisbook,Grapevine:
Everybodytalks aboutproducts s and ser-
vices, and they talk about them all the
time.WordofmouthisNOTaboutiden-
tifyinga small subgroup ofhighlyinflu-
entialorwell-connectedpeopletotalkupa
productorservice.It’snotaboutmavens
orbeesorcelebritiesorpeoplewithspecial-
istknowledge.It’sabouteverybody.(Bal-
terandButman,2005).
ColumbiaUniversitySociology Profes-
sor Duncan Watts agrees, arguing what
he calls the “influentials hypothesis” is
based on untested assumptions and in
mostcasesdoesnotmatchhowdiffusion
operates in the real world. He e observes
TABLE1
NewMediaUsage
PleaseIndicateHowFrequentlyYouPerform
theFollowingActivities
Percent“VeryFrequently”
or“Frequently”
.............................................................................................................................................................
Forwardinformationfoundontheinternetto
colleagues,peers,family,orfriends
59%
.............................................................................................................................................................
Readnewspapersonline
48%
.............................................................................................................................................................
Readmagazinesonline
25%
.............................................................................................................................................................
Readablog
24%
.............................................................................................................................................................
Listentoradiofeedsviatheinternet
23%
.............................................................................................................................................................
Participateinanonlinecommunity,suchas
myspace.comorfriendster.com
22%
.............................................................................................................................................................
UsePVRtechnology,suchaTiVoorDVR
22%
.............................................................................................................................................................
Vieworpostvideosonawebsite,suchasyoutube.com
17%
.............................................................................................................................................................
Listentosatelliteradio
16%
.............................................................................................................................................................
Createorparticipateinablog
13%
.............................................................................................................................................................
Subscribetoapodcast
6%
.............................................................................................................................................................
Createapodcast
2%
.............................................................................................................................................................
Source:HarrisInteractive AnnualRQsm.Base=6,205U.S.adults(18+)familiarwithone ormoreofthe10“most
visible”U.S.companies.
WORD-OF-MOUTHRESEARCH
December 2007
JOURNALOFADVERTISINGRESEARCH
399
RasterEdge .NET Document Imaging Trial Package Download Link.
View & Process. XImage.Raster. Adobe PDF. XDoc.PDF. Scanning. XImage.OCR. Microsoft Office. View & Process. XImage.Raster. Adobe PDF. XDoc.PDF. Scanning. XImage.
add hyperlink to pdf online; add page number to pdf hyperlink
C# PDF Library SDK to view, edit, convert, process PDF file for C#
RasterEdge PDF SDK for .NET package offers robust APIs for editing PDF document hyperlink (url) and quick navigation link in PDF bookmark.
add links to pdf in preview; add link to pdf
that“mostsocialchangeisdrivennotby
influentials,butbyeasilyinfluencedindi-
vidualsinfluencingothereasilyinfluenced
individuals”(Wa�sandDodds,2007).
We have observed that those e who are
most influential l in a given category y are
o�ennot thosewhomyouwouldexpect.
Asto special roles, we e have found that
mavens,connectors,andsalesmenarenot
usuallyseparate individualsasGladwell
suggests.Rather, they y are traitsthat can
existseparately orintandem(invarious
degrees)inthesameindividual.Further-
more,any particular individual may as-
sumeadifferentrole(giverorreceiverof
WOM)inthe social network,depending
onthetopicunderconsideration.
Infact,eachofusbelongstomultiple
socialnetworks.Thepeopletowhomwe
talkaboutautomobilesarenotnecessarily
theonestowhomwetalkaboutlaundry
soap. Inaddition,the e size and composi-
tionofoursocialnetworkvaryfromone
categorytothenext.Somearelarger,with
most people participating, while others
are more specialized.What isimportant
isto understand howthe specific c social
networkinyourcategoryoperates,andin
particular,whichindividualswithinthat
Figure1 WordofMouth:ATwo-WayExchange
Whatisimportantistounderstandhowthespecificsocial
networkinyourcategoryoperates,andinparticular,which
individuals within that social network are most active
in creating and spreading messages about your product
categorytoothers.
WORD-OF-MOUTHRESEARCH
400
December 2007
JOURNAL OF ADVERTISING RESEARCH
Synthesis Alliance/
How to C#: Basic SDK Concept of XDoc.PDF for .NET
XDoc.PDF for .NET allows C# developers to edit hyperlink of PDF document, including editing PDF url links and quick navigation link in bookmark/outline.
pdf hyperlink; pdf link to attached file
VB.NET PDF: Basic SDK Concept of XDoc.PDF
XDoc.PDF for .NET allows VB.NET developers to edit hyperlink of PDF document, including editing PDF url links and quick navigation link in bookmark/outline.
convert excel to pdf with hyperlinks; add hyperlink pdf file
socialnetworkaremostactiveincreating
andspreadingmessagesaboutyourprod-
uctcategorytoothers.
The results inFigure 1illustrate e how
theproportionof
thepopulationinvolved
inWOMvariesfromonecategorytoan-
other. More e of ustalk k about restaurants
(94 percent) and computers (94percent)
thanaboutpersonalcareproducts(65per-
cent)orathleticshoes(45percent).Acloser
lookat
thedatashowsthat
there isalot
ofoverlap;the vastmajority of
informa-
tionprovidersare also informationseek-
ers, and vice versa. That is how social
networksoperate:wegatheropinionsfrom
others,weincorporatethemintowhatwe
knowandfeel,andwepassthatalongto
others.
Thisstudy alsoconfirmsthat virtually
everybody participates in one or more
socialnetworks.Infact,only1percentof
respondentssaidthey donot participate
atallinprovidingorseekinginformation
inanyofour14measuredcategories.
Anotherinterestingobservationisthat
the“supplyanddemand”of
information
andadvicevariesbycategory,whichmay
giveuscluesastothenatureofWOMin
thatcategory.Ifwelookat
themiddleof
the chart, 18percent of respondentssay
they seek informationonfinancial prod-
uctsandservicesto a great extent 
,
yetonly
8percentprovidefinancialinformationto
a  great  extent,  suggesting  this  is  a  topic
where people e are more e likely to turnto
experts. Conversely, 15 percent actively
provideinformationaboutpolitics,whereas
only10percentactivelyseekit,sopolitics
isatopicwheresomepeopleshareinfor-
mationevenwhenothersare notasking.
Active providers also outnumber active
seekers,althoughtoalesserextent,when
it comes to o movies, personal care prod-
ucts,andcompanies.
By zeroing inonthe most active par-
ticipantsinthesocialnetwork,thosewho
seekorprovideinformationandadviceto
agreatextent, we can learn much:
1. Weseesignificantdifferencesinsocial
networkactivitybasedondemographic
traits,suchasage andgender.Forex-
ample,themajorityof
thosewhopro-
vide to a great extent on vehicles, financial
services,computers,andpoliticsaremen,
whileactiveprovidersonpersonalcare,
OTCmedications,nutrition,andhealth
care providers are e more likely to be
women. Gender r differences, interest-
ingly,aremuchlesspronouncedwhen
itcomestoseekinginformation.
2. There is a lot more e social network/
WOMactivityinsomecategoriesthan
others.HarrisInteractive’sSocialNet-
work
Commerce
IndexTM (Synthesis/
Harris,2006)providesanother useful
waytoquantifythelevelofWOMac-
tivitywithinagivencategory.SeeFig-
ure2.Byfocusingonlyonthelevelof
activeseekingandproviding,theindex
portraysthemultipliereffectofWOM
andkeepsourfocusonthosemostlikely
tohaveanimpactonthespreadofWOM.
Theindexisalsoausefulmeansfor
comparingtherelativeinvolvementof
different subgroupsinWOMwithina
givencategory.Forexample,whencal-
culatingtheindexscoresfor“nutrition
andhealthyeating,”wecanseevarying
degreesofactivityamong agegroups,
educationala�ainmentgroups,income
groups,and householdsizes(see Fig-
ure3).Thiscanhelpmarketerstopri-
oritizetheiractivitiesandsegment
their
targetaudiencestomakesuretheyare
reachingthosemostlikelytoengagein
WOM.
3. The notion that
there is amonolithic
block of cross-category influentials
Figure2 Social NetworkCommerce Index
TM
WORD-OF-MOUTHRESEARCH
December 2007
401
Our research bears this out. In a study 
conducted for Synthesis Alliance (Synthesis/ 
Harris 2006), U.S. adults were asked to 
characterize the extent to which they seek 
or use information and advice from other 
people, as well as the extent to which they 
offer or are asked to provide information and 
advice to other people—on 14 different 
categories, including goods, services, and 
intangibles. 
JOURNAL OF ADVERTISING RESEARCH
Synthesis/Harris
VB.NET Word: How to Process MS Word in VB.NET Library in .NET
Besides, here is the quick link for how to process Word document within We are dedicated to provide powerful & profession imaging controls, PDF document, image
add url to pdf; clickable links in pdf files
comprising10 percent t of the e popula-
tionisnotsupportedbyourdata.While
itisnotunusualforpeopletobeasource
ofWOMforseveralcategories,asshown
below,the10percentwhoaremostac-
tive inprovidinginformationonly do
soinaboutfivecategories.Weturnto
differentpeopleforinformationondif-
ferent topics.(SeeFigure4).
There doesappear to o be a handful of
true “market mavens” who provide in-
formationandadvicetoagreatextenton
all 14 of the categories measured, but
theynumberfewerthan2percent ofre-
spondents. Most of us rarely encounter
this kind ofperson—our day-to-day y in-
teractionsarewithaverageconsumerslike
ourselves. Rather than spend resources
trying to findandtarget these e supposed
influentials,marketersshouldworktoun-
derstandwhohasthegreatestimpacton
the spread of WOM in their particular
category and figure out ways to give
themapositiveexperiencewiththebrand,
so they will l be more likely to passthat
along.
PRINCIPLE#2
Word-of-mouthhappensinthecontextofa
specific situationandoccasion.
WOMhasanumberofdimensionsthataf-
fecthowitspreadsforyourparticularcat-
egoryorbrand.For example,howmany
people does anindividualcommunicate
withabout thetopic?Howfrequently?How
relevant is themessage e tothem person-
ally?Howaccurateistheinformationthat
ispassedalong?Arewetalkingaboutpos-
itiveornegativemessages?Theseandother
dimensionsdeterminewhetherWOMwill
spreadquicklyorslowly,toabroadgroup
ortoanarrowgroup,ornotatall.Toun-
derstandhowWOMworks,weneedtoac-
count for thesedifferentdimensionsand
howtheyareinterconnected.
Figure5illustratessomeof thekeydi-
mensionswefocusonaswedoWOMre-
other.InanalyzingWOMforourclients,
wetrytounderstandasmanyof thesedi-
mensionsaspossiblesoastoproducein-
sightsthatarebothaccurateandactionable.
Current WOM research has given us
ways to operationalize many of these di-
mensions.Forexample:
• UnderA�ributes of the Source,we e look
at credibility and persuasiveness of the
personprovidingthemessage,because
these affect whether or not the message
will be acted on or passed along to
others.
• UnderRateofActivity,wemeasurehow
likelysomeoneistoactivate thesocial
network (either to seekor providein-
formation),andhowquicklyandhow
o�en  they  share  opinions  about  the
product orserviceunderstudy.
• Under Personal Relevance,we consider
therationalandemotionalcomponents
of the e message,using ourexpertise in
Means-Ends theory research, which
teachesusthatvalue-ladenemotionally-
charged appeals are much more e rele-
vantandthereforepersuasive.
• UnderPolarity,welookatwhetherthe
tone of the communication is negative
Figure3 Social Network Commerce Index
TM
for “Nutrition
andHealthyEating” amongDemographic Subgroups
Figure4 Few Influence a
Large Number ofCategories
WORD-OF-MOUTHRESEARCH
402
December 2007
Synthesis/Harris
Synthesis/Harris
JOURNAL OF ADVERTISING RESEARCH
search, and the fact that they all affect each
VB.NET PDF: Create PDF Document Viewer in C#.NET for Document
reading PDF document in ASP.NET web, .NET Windows Forms and mobile developing applications respectively. For more information on them, just click the link and
add hyperlink to pdf; check links in pdf
VB.NET Word: VB Code to Create Word Mobile Viewer with .NET Doc
For the respective tutorials of these Document or Image Mobile Viewer in VB.NET prorgam, please link to see: PDF Document Mobile Viewer within VB.NET
adding links to pdf document; pdf edit hyperlink
orpositive,andwhetheritcontributes
towhatwecalltheperceptualequities
ortheperceptualdisequitiesofthesub-
ject brand/product/company.
Butwithotherofthesedimensions,we
arejustbeginningtounderstandhowthey
influenceWOMactivitywithinsocialnet-
works,andtheresultingimpactsoncon-
sumerbehavior.Webelievethatthefuture
of WOM research will largely focus on
finding ways to o better understandthese
dimensions,developing reliable methods
tomeasurethem,andaccumulatingexpe-
rience in translating those insights into
actionimperativesforbusinessandmar-
ketingdecisionmakers.
Inadditiontoongoingeffortstobetter
measure the above constructs, we have
startedtodevelopalternativeapproaches
to modeling their complex interactions.
OnesuchformulationisshowninFigure6.
• The probability of a selected consumer
holding a high, medium, or low per-
ception of the reputation of a firm is
conditionally dependent on the influ-
enceandcredibilityofaset ofinforma-
tion channels that impact the e various
dimensions of the e reputation quotient
(RQ).
• Eachinformationchannel hasadirect
impact(high,medium,orlow)oneach
of the e six x dimensionofreputation.In
turn,that levelofimpactisdependent
ontheprobabilityofaconsumerbeliev-
ingwhether or not thatinformationis
credible.
• Sensitivity analyses have beenrunon
reputation for each of four sample
companies:
relative impact of each information
channelonWOM,
relative impact of each information
channel on each of the six dimen-
sionsofreputation,and
impactofeachreputationdimension
onoverallreputation.
Intheexampleshownattheendofthis
article(seeFigure7),weapplythismodel
to a proprietary data set from the RQ
study (Harris Interactive, 2006b) using
BayesianBeliefNetworks.
PRINCIPLE#3
Peoplemakedecisionsbasedonacomplex
interplayofcognitivepreferencesandemo-
tionalbenefits.
Wehaveatremendousamountofexperi-
ence showing that human beings make
decisionsabout productsbasedonthree
levels:(1)theattributesofaproduct,(2)
thefunctionalbenefitsandemotionalcon-
sequencesderived from those attributes,
and (3) the personal values that those
consequencesreinforce.Valuesarebydef-
initiondeeplyemotional,highlypersonal,
andpowerfullymotivating.Thebestway
topersuadesomeone todo somethingis
toappealtothevaluesthatmatterdeeply
to them. When a message succeeds in
doing that, we say it has high personal
relevance.
For example,Ibuy BrandXbecauseit
cleansbetter,whichleavesfewergermsin
my house, so o my kidswill not get t sick,
whichmakes me abetter r parent. Ifyou
wanttosellmeBrandX,don’tjusttellme
Figure5 Dimensions ofaSocial Network
WORD-OF-MOUTHRESEARCH
December 2007
JOURNALOFADVERTISINGRESEARCH
403
it cleans better.Show me why y I should
care about that,by linkingthat attribute
inacrediblewaytotheconsequencesand
values that make me tick—in this case,
mydeepdesiretobeagoodparent.Now
you’vegotmyattention.
There is both anart and a science e to
crafting these kinds of communications
that “persuade by reason and motivate
through emotion.” If we do our home-
work (which involves well-designed re-
searchintoconsumermotivationsandthe
decision-making processas it t appliesto
your particular category), we e can maxi-
mizepersonalrelevanceonboththecog-
nitiveandemotionaldimensions.
This values-based approach provides
deepstrategicinsightsfordevelopingmar-
ketingcommunicationsthatproducemea-
surableresultsintherealworld.Overthe
pastdecade,sixnationaladvertisingcam-
paignsonwhichwehaveworked,where
values research provided the strategic
framework,havewontheAdvertisingRe-
search Foundation’s David Ogilvy Re-
searchAward,whichrecognizesthe role
of research in contributing to successful
advertising campaigns,asjudgedbyde-
monstrablein-marketresults.
While personal relevance is afounda-
tionofany type ofmarketing,itisespe-
ciallyimportantwhentalkingaboutWOM.
Themorepersonallyrelevantourproduct
and our messages are, the more likely
consumersare toengage withthe prod-
uct, andmore e likely to pass along g mes-
sages to others. When you know what
emotionalchordsyourproductandyour
messagesaretouchingwithinyouraudi-
ence,youcanappealinsubtle but pow-
erfulways,buildingloyalty,while atthe
sametimefacilitatingsharing.
For example, Hard Rock Café knows
theirpatronstakeprideinbraggingabout
having eatenat the e restaurant infar off
places.Part oftheirdeliberatestrategyis
toappeal to thatsense ofbelonging to a
specialgroup.Therefore,theysellT-shirts
so you can come e back and showevery-
body you ate at the e Hard Rock Café in
Tokyo.Not only are they reinforcingthe
senseofpride,theyarefacilitatingWOM
because others will see the T-shirt and
naturally strike up aconversationabout
theirexperiences.
PRINCIPLE#4
Theconsumerenvironmentinwhichword
ofmouthtakesplaceisconstantlychanging.
Figure6 RQ
SM
WOMModel Structure
WORD-OF-MOUTHRESEARCH
404
JOURNALOFADVERTISING RESEARCH
December 2007
Earlierwedescribedtheconsumermarket-
placeasacomplex,dynamicsystem.WOM,
moresothanothertypesofinformationex-
change in the marketplace,can change
quickly.Becauseofthat,itisimportantto
continuallymeasureandmonitorwhatis
goingon,soyoucanspotdevelopmentsas
theyoccur,andquicklymakecoursecor-
rectionsortakeotheractionsthatwillpos-
itivelyinfluencethesystem.
For example,manycompaniesandor-
ganizationsroutinelymonitorblogsabout
their products,services,andreputation.
More thanafewemploy full-time blog-
gers,who not only report on negative
WOM,butactivelyparticipatebyposting
messagestocorrectfactsandcountermis-
perceptionsthatarise.Theseprofessional
“blogmonitors”donottrytohidethecom-
panyforwhomtheywork.Thistranspar-
ency isvitaltokeepthiskindofactivity
frombackfiring.Wehavefoundthataccu-
rate,unassailablefactscantrumppeople’s
normalskepticismtowardinformationthat
comesfromthecompany.Thisisjustone
exampleofhowmarketerscanbecomepart
ofthedynamicexchangeofideas,rather
thanjusthelplessobserversonthesidelines.
PRINCIPLE#5
Thediffusionandimpactofmessageswithin
thesocialnetworkvariesbasedonthepo-
larity (positive/negative) of the messages
beingcommunicated.
Yourbusinessplanshouldencompassnot
onlyhowtoinfluenceandleverageposi-
tive WOM, but also how to neutralize
negative WOM. Recognizing the differ-
ences between how the social network
deals with positive and negative mes-
sagesisimportant.
Ingeneral,weknowthatnegativemes-
sagestendtospreadmorequicklywithin
asocialnetwork.Forexample,studiesby
Burson-Marstellerhavefoundthat“tech-
fluentials”will passalong positive mes-
sagestoanaverageof13people,butthey
willsharenegativemessageswithanav-
erageof17people(DeitzandÇakim,2005).
The reasonspeople choose whether or
nottopassalonganegativemessageare
entirely different from those that influ-
encethemwhenthemessageispositive.
AWirthlinWorldwidestudypublishedin
2004 showed that email l users are most
likely to forward negative news about
financial fraud, health, or safety, all of
whichhavehighpersonalrelevance.Con-
sumersseemtobemotivatednotbyspite,
but rather by a genuine desire to save
othersfrommakingbaddecisions(Wirth-
linWorldwide, 2004). Marketers can use
thatinsighttotheiradvantageincombat-
ingnegativeWOM.
There are different types of negative
WOM, depending on how it originates,
eachofwhichrequiresadifferentkindof
response:
• WherenegativeWOMarisesfromdis-
satisfied customers, negative reviews
or products that fail to meet t expecta-
tions, work to fix the problems and
improveproducts.
• NegativeWOMsparkedbyattacksfrom
critics or r competitors usually y contains
alarmism, half-truths,or r outright lies.
Today’stechnologyandthemedia’sap-
petiteforcontroversygivethesedetrac-
torsabiggerspotlightthantheydeserve.
Herewewanttorespondaggressively
andget outthefactstotrytoturnthe
tideofnegativeWOM.
• WOMalsospreadsrapidlywhenthere
isanunexpectedproductfailure,safety
issue, or scandal that has its basis in
truth,butmaybeblownoutofpropor-
tion.Crisismanagement isatopicfor
another article, but again, honest re-
sponseandquickactiontofixtheprob-
lemareparamount.
Asacompanycomestounderstandthe
components of the social network sur-
rounding its products and brands, and
establishesmechanismsto “listenin”on
WOM, it will be in a better r position to
respond quickly, specifically, and can-
didlytonegativeWOM,minimizingharm
tosalesandreputation,andoftenenhanc-
ingitsimageintheprocessbecausecon-
sumersgivecompaniesandorganizations
creditforhonestlyhandlingproblems.
PRACTICALAPPLICATIONSOF
WORD-OF-MOUTHRESEARCH:
ANEXAMPLE
Anexampleofthepracticalapplicationof
WOM research can be found in Harris
Interactive’sAnnual RQSM study y (Harris
Interactive,2006b),alreadycited.
Alongwithrankingsoncomponentsof
reputation,thestudyincludesaseriesof
questions about mediausage and social
networkactivity,toshedlightontherel-
ativeimpactofWOMversusothersources
of information in formulating the opin-
ions that drive the reputation of these
companies.While thisis not t as compre-
hensiveananalysisascouldbedonefor
anysinglecompany,itdoesillustratesome
of the key dimensions of WOM (de-
scribedearlyunderPrinciple#2),andhow
we approach measuring g those to o gain a
Whilepersonalrelevanceis a foundation of anytype of
marketing,it isespeciallyimportant whentalkingabout
WOM.
WORD-OF-MOUTHRESEARCH
December 2007
JOURNALOFADVERTISINGRESEARCH
405
fuller understanding, inthiscase,ofthe
driversofcorporatereputation.
UsingdatafromtheRQ,andthemodel
showninFigure6,webuiltaseriesofBayes-
ianBeliefNetworks,one foreachoffour
examplefirms.Herearejustafewofthe
thingswe learned,whichreinforce some
ofthefiveprincipleswehavediscussed:
1. WOM plays a significant role—often
morethananyothersource—ininflu-
encingperceptions,yetitssignificance
variesfrom one dimensionofreputa-
tionto another andfrom companyto
company.Ofthesixdimensionsofrep-
utation we e measured,emotional appeal
(trust,goodfeelings,andrespect)con-
sistentlyhasthestrongestinfluenceon
corporatereputation,followedbyper-
ceptionsaboutthecompany’sproducts
andservices.Those twodimensions,in
turn,aredrivenheavilybyWOM,which
accountsforasmuchashalfofeachdi-
mension.(SeeFigure7.)Amongthefour
examplecompaniesanalyzed,theonly
exceptionsarecompaniesCandD,where
emotionalappealisdrivenmostlybyper-
sonalexperience,andcompanyC,where
perceptionsaboutproducts andservices
aredrivenmainlybyadvertising.
2. As shown above, WOM has a strong
influencealongbothrationalandemo-
tionaldimensions(Principle #3).Prod-
ucts andservices isarational attribute,
while emotional appealgetsat personal
emotionsandhigher-levelvalues.
3. The probability of a given consumer
holding a high,medium, or lowper-
ception of the e reputationofa firm m is
conditionally dependent on the e influ-
ence and credibility of the six infor-
mation channels measured. Next to
personalexperience,WOMisthesource
withthehighestpositiveinfluenceand
credibility.(SeeTable 2.)
4. Whilewehavespokenabouttheinflu-
enceofWOM,itisitselfinfluencedby
othersourcesofcommunication.Inthat
sense,WOMactsasasortofmultiplier.
AsshowninFigure6,eachinformation
channelhasadirectimpactoneachof
Figure7 InfluenceofWOMonCompanyReputation
WORD-OF-MOUTHRESEARCH
406
JOURNALOFADVERTISING RESEARCH
December 2007
the six dimension of reputation, but
each also o has anindirectinfluence by
coloringtheWOMthatisreceivedand
passedalong.
Our analysis allows us to quantify
this indirect influence.As shown be-
low,personalexperiencehasthegreat-
estinfluenceonWOM,butothersources
are also significant.For example,me-
diastoriesaboutcompanyAhadnearly
as much influence on WOM as per-
sonalexperience.(SeeTable3.)
5.Polarity(negativeversuspositive)ofthe
WOMmessage issignificant, as dis-
cussedinPrinciple#5.AspartoftheRQ
study,respondentswereaskedtochar-
acterize their r WOMcommunications
about eachcompany on ascale e from
“verypositive”to“verynegative.”From
2005to 2006,we sawasignificant in-
creaseincompany C’soverallreputa-
tionscore,whilecompanyA’sscorewent
down.Acloserlookatthedatasuggests
thatthetoneofconversationaboutthese
twocompaniesmaybeplayingastrong
roleindrivingtheperceptionsleading
tothisoutcome:ForcompanyC,thera-
tio ofvery positive to o very negative
WOM rosefrom 4:1in2005to 61
2
_:1in
2006.Overthe same period,company
A’sverypositivetoverynegativeratio
fellfrom5:1downto1:1.
UNDERSTANDINGWORD-OF-MOUTH
USINGCOMPLEXITYSCIENCE
TofullyunderstandWOM,withallofits
movingparts,requiresasophisticatedan-
alyticalapproachthatisbeyondthereach
of traditional marketing research meth-
ods.However,anewgenerationoftools
nowmakesthispossible.
Inthe mid1980s,ateam ofinterdisci-
plinaryscientistsformedtheSantaFeIn-
stituteandsetouttodevelopatheoretical
framework to o describe complex systems
made up p of multiple e interconnected ele-
ments.Drawingonsystemstheory,cyber-
netics,chaostheory,andneuralnetworks,
theireffortsculminatedinthefundamen-
talconceptsofComplexityScience:(1)ev-
erything is related,(2) nothing g islinear,
and (3) small changes can create un-
expectedanddisproportionateoutcomes.
Theseconceptsareusedtoconstructcom-
puterized models of complex systems.
Ratherthanrelyingonlinearstatistics,Com-
plexityScienceusesanapproachknownas
agent-basedmodeling(ABM).ABMmod-
elswerefirstdesignedtounderstandphe-
nomenasuchashivesofbees,flocksofbirds,
ortrafficjams,where the observable ac-
tionsofthegroupemerge,oftenunpredict-
ably,fromtheinteractionsofitsindividual
members.The flexibilityofsuchmodels,
andthefactthattheycanbeadjustedto
accountforawidevarietyoffactors,make
themidealfordescribinghowWOMworks.
TABLE2
WordofMouthSeenas Influential,CredibleSource
InformationSource
MyPerceptionsare
Influencedtoa
PositiveExtent
byThisSource
ThisisaVery
CredibleSource
.............................................................................................................................................................
WOM
81%
34%
.............................................................................................................................................................
Advertisingforthecompany
77%
16%
.............................................................................................................................................................
Companypublicrelationsactivities
67%
15%
.............................................................................................................................................................
Myownpersonalexperiencewiththecompany 85%
70%
.............................................................................................................................................................
Opinionsofthecompany’semployees
66%
33%
.............................................................................................................................................................
Mediastoriesaboutthecompany
68%
15%
.............................................................................................................................................................
Source:HarrisInteractive AnnualRQ
SM
.OnlinesurveyconductedSeptember21
October23,2006.Base=6,205U.S.
adultswhoratedthetop10companies.
TABLE3
InformationChannel InfluenceonWordofMouth
%Impact
............................................................................................................................
Channel
CompanyA
CompanyB
CompanyC
CompanyD
.............................................................................................................................................................
Advertisementsforcompany 9%
10%
13%
25%
.............................................................................................................................................................
Employeeopinion
12%
14%
13%
26%
.............................................................................................................................................................
Mediastories
34%
12%
14%
2%
.............................................................................................................................................................
Personalexperience
36%
43%
38%
37%
.............................................................................................................................................................
Publicrelations
10%
20%
22%
10%
.............................................................................................................................................................
Source:HarrisInteractive AnnualRQ
SM
.OnlinesurveyconductedSeptember21
October23,2006.Base=U.S.adults
(18+)familiarwithcompany.Actualresults,companynamesblinded:
CompanyA=automotiveindustry,n=649respondents
CompanyB=retailindustry,n=616respondents
CompanyC=ITindustry,n=663respondents
CompanyD=electronicsindustry,n=698respondents
WORD-OF-MOUTHRESEARCH
December 2007
JOURNALOFADVERTISINGRESEARCH
407
Documents you may be interested
Documents you may be interested