open pdf file in asp.net using c# : Add text box in pdf application Library tool html asp.net wpf online 201210171029554160-part637

MANAGEMENT SCIENCE
Vol.54,No.3,March2008,pp.477–491
issn0025-1909eissn1526-55010854030477
inf
orms
®
doi10.1287/mnsc.1070.0810
©2008INFORMS
Online Consumer Review: Word-of-Mouth as a
New Element of Marketing Communication Mix
YuboChen
EllerCollegeofManagement,UniversityofArizona,Tucson,Arizona85721,
yubochen@eller.arizona.edu
JinhongXie
WarringtonCollegeofBusiness,UniversityofFlorida,Gainesville,Florida32611,
jinhong.xie@cba.ufl.edu
A
s a new type of word-of-mouth information, online consumer product review is an emerging market
phenomenonthat isplayinganincreasingly importantroleinconsumers’purchasedecisions.Thispaper
arguesthat onlineconsumerreview,atypeofproduct informationcreatedbyusersbasedonpersonalusage
experience,canserveasanewelementinthemarketingcommunicationsmixandworkasfree“salesassistants”
tohelpconsumersidentifytheproductsthatbestmatchtheiridiosyncraticusageconditions.
Thispaper developsa normative e modelto addressseveralimportant strategicissuesrelated toconsumer
reviews.First,we showwhen andhowthe seller shouldadjust itsownmarketingcommunicationstrategy in
responsetoconsumerreviews.Ourresultsrevealthatifthereviewinformationissufficientlyinformative,the
twotypesofproductinformation,i.e.,theseller-createdproductattributeinformationandbuyer-createdreview
information,willinteractwitheachother.Forexample,whentheproductcostislowand/ortherearesufficient
expert (more sophisticated)product users,thetwotypesofinformationare complements,andtheseller’sbest
responseistoincreasetheamountofproductattributeinformationconveyedviaitsmarketingcommunications
afterthereviewsbecomeavailable.However,whentheproductcostishighandtherearesufficientnovice(less
sophisticated)product users,thetwo typesofinformationare substitutes,and theseller’sbest t responseisto
reducetheamountofproductattributeinformationitoffers,evenifitiscost-freetoprovidesuchinformation.
Wealsoderivepreciseconditionsunderwhichthesellercanincreaseitsprofitbyadoptingaproactivestrategy,
i.e., adjusting its marketing strategies even before e consumer reviews become available. Second, we identify
product/market conditions under which the e seller benefits from facilitating g suchbuyer-created information
(e.g.,byallowingconsumerstopostuser-basedproductreviewsontheseller’swebsite).Finally,weillustrate
theimportanceofthetimingoftheintroductionofconsumerreviewsavailableasastrategicvariableandshow
thatdelayingtheavailabilityofconsumerreviewsforagivenproductcanbebeneficialifthenumberofexpert
(moresophisticated)productusersisrelativelylargeandcostoftheproductislow.
Keywords: onlineconsumerreview;word-of-mouth;productreviewinformation;marketingcommunications;
socialinteractions
History: AcceptedbyJagmohanS.Raju,marketing;receivedJune21,2005.Thispaperwaswiththeauthors
11
1
2
monthsfor2revisions.
1. Introduction
The Internet and information technology provide a
newopportunityforconsumerstosharetheirproduct
evaluations online (Avery et al. 1999). Amazon.com
began offering consumers an option to post their
commentson products on its website in 1995. Cur-
rently, Amazon.com has about 10million consumer
reviews on all its product categories, and these
reviews are regarded as one of the most popular
and successful featuresofAmazon (New York Times
2004).Inrecentyears,anincreasingnumberofonline
sellers (e.g., BevMo.com, BN.com, cduniverse.com,
circuitcity.com, GameStop.com, computer4sure.com,
c-source.com, half.com, goodguys.com, wine.com)
have adoptedasimilarstrategy. These onlinesellers
inviteusersoftheirproductsto postpersonalprod-
uct evaluations on the sellers’ websites or provide
their customers with consumer review information
offered by some third-party sources such as Epin-
ions.com. Online consumer reviewsare commonfor
manyproduct categories such asbooks, electronics,
games,videos,music,beverages,andwine.
Recent evidence suggests that consumer reviews
have becomeveryimportant for consumer purchase
decisionsandproductsales.AstudybyForresterRe-
searchfindsthathalfofthosewhovisitedtheretailer
siteswithconsumerpostingsreportedthatconsumer
reviewsareimportantorextremelyimportantintheir
477
Add text box in pdf - insert text into PDF content in C#.net, ASP.NET, MVC, Ajax, WinForms, WPF
XDoc.PDF for .NET, providing C# demo code for inserting text to PDF file
adding text to a pdf document; adding text to a pdf in preview
Add text box in pdf - VB.NET PDF insert text library: insert text into PDF content in vb.net, ASP.NET, MVC, Ajax, WinForms, WPF
Providing Demo Code for Adding and Inserting Text to PDF File Page in VB.NET Program
add text boxes to pdf; add text fields to pdf
ChenandXie: OnlineConsumerReview:Word-of-MouthasaNewElementofMarketingCommunicationMix
478
ManagementScience54(3),pp.477–491,©2008INFORMS
buyingdecisions(Los Angeles Times 1999). Based on
the data from Amazon.com and BN.com, Chevalier
and Mayzlin (2006) find that online book reviews
have a significant impact on book sales. Liu (2006)
shows that consumer reviews at the Yahoo Movies
websitehasasignificanteffectonboxofficerevenue.
However, not all online sellers supply consumer
reviewsontheirwebsites.Forexample,ChenandXie
(2004)examine three product categories: MP3 play-
ers,PDAs,andvideogames.Theyidentifyalistof68
onlinesellersfromthereferrallistoftheleadingshop-
pingagent mySimon.cominJune18,2003, and find
that46outof68onlinesellersdidnotofferconsumer
reviews.
Online consumer review is a new product infor-
mationchannelwithgrowingpopularityandimpor-
tance. It has generated considerable attention in
practitioners and popular presses. Sellers face vari-
ousimportantstrategicdecisionsregardingconsumer
review information. For example, when consumer
reviewsappear,shouldaselleradjustitsowncommu-
nicationstrategytobestrespondtosuchaconsumer-
created informationchannel, and how?Under what
conditionsdoesthesellerbenefitfromfacilitatingthe
creationanddisseminationofsuchuser-basedreview
informationbyallowingconsumerstoposttheircom-
ments on its own website (e.g., Amazon.com)? To
better understand the e fundamental role of thisnew
informationchannelinthemarketplaceanditsstrate-
gicimplicationsto online marketers, moreacademic
researchisurgentlyneeded.
Severalrecentstudieshavebeguntoexamineonline
consumer-created information from the perspective
ofinformationcredibility.Consumer-createdinforma-
tion islikelyto be more credible thanseller-created
informationbecausecredibilityofinformationisoften
positivelyrelatedtothetrustworthinessofthe infor-
mationsource(WilsonandSherrell1993).Dellarocas
(2003)reviewsthe relationship between online con-
sumerfeedbackinformationandanunknownseller’s
reputation. Mayzlin (2006) studies the credibility of
the promotionalmessagesin online e chat roomsand
theimplicationofsuchnewinformationchannelson
sellers’profitability. Furthermore, some recent stud-
ies(FayandXie2008,XieandGestner2007)suggest
thatconsumer-createdinformationallowsthesellerto
implementsomemarketingstrategiesthatmaynotbe
credible otherwise e (e.g., probabilistic selling, service
cancellation).Thesestudieshaveadvancedourunder-
standingofconsumer-createdinformation.Animpor-
tant but underexplored aspect of consumer reviews
istheir degree ofrelevance to consumers. We argue
that online e consumer reviews can be e deployed as a
new element inthe marketingcommunications mix
andworkasanonlineseller’sfree“salesassistants”
(Wernerfelt1994a)tohelpconsumerstoidentifyprod-
uctsthatbestmatchtheirneeds.
Toexaminesuchamatchingfunctionofonlinecon-
sumerreviews,wefirstpresentanempiricalstudyto
illustratehowthisemerginginformationsourceisdif-
ferent fromothertypesofproductinformation,such
as third-party product reviews. We then develop a
normativemodeltoaddressseveralspecificquestions
regarding a firm’s strategic decisions vis-à-vis con-
sumerreviews.
Our empirical study suggests that, different from
third-partyproduct reviewsthat emphasizetheper-
formance of a product based on its technical spec-
ifications, consumer reviews tend to examine the
performance of a product from the perspective of
itsabilityto matchthe consumers’ownusagesitua-
tions.Ourstrategicanalysisrevealsseveralimportant
findings. First, weshowthat the two typesofinfor-
mation—consumerreviewsandseller-createdproduct
attribute information—can be complements or substi-
tutes. Suchinteractionexists whenthe e review infor-
mation is sufficiently informative. The direction of
the interaction (i.e., complementary or substitutive)
is determined by the characteristics of the product
and market. When the product cost is low and/or
therearesufficientexpert (moresophisticated)prod-
uct users, the e two types ofproduct informationare
complements.Inthiscase,thesellershouldincreasethe
amountofitsownproductattributeinformationcon-
veyedtopotentialcustomerswhenconsumerreviews
become available. When the product cost is high
and there are sufficient novice (less sophisticated)
product users, the two types of product informa-
tionaresubstitutes.Here,thesellershoulddecreaseits
productattributeinformationsupplywhenconsumer
reviewsbecomeavailable.Inaddition,we showthat
ifthesellercananticipatetheavailabilityofconsumer
reviews, it is possible to adopt a proactive strategy
byadjustingitsmarketingstrategiesevenbeforecon-
sumerreviewsbecomeavailable.Second,ouranalysis
reveals that allowing consumers to post user-based
product reviewsonthe seller’swebsite can increase
ordecreaseprofitdependingonproduct/marketcon-
ditions. We show thatit isdetrimentalto aseller to
supplyconsumerreviewsunlesssuchinformationis
sufficientlyinformative. We also find thatsupplying
online consumer reviews is more likely to be ben-
eficial to the seller when there e are sufficient novice
consumers (e.g., for technology-intensive products).
Finally,ourresultsrevealthat ifitispossibleforthe
sellertodecidewhentoofferconsumerreviewsatthe
individualproductlevel, it maynot alwaysbeopti-
maltoofferthemataveryearlystageofnewprod-
uct introduction, even ifsuch reviewsare e available.
Delayingthe availabilityof consumer reviews for a
givenproductcanbebeneficialifthenumberofthe
VB.NET PDF Text Box Edit Library: add, delete, update PDF text box
Barcode Read. Barcode Create. OCR. Twain. Add Text Box. |. Home ›› XDoc.PDF ›› VB.NET PDF: Add Text Box. VB.NET PDF - Add Text Box to PDF Page in VB.NET.
add text pdf professional; how to add text to pdf file
C# PDF Text Box Edit Library: add, delete, update PDF text box in
C# PDF: Add Text Box. C#.NET PDF SDK - Add Text Box to PDF Page in C#.NET. C# Explanation to How to Add Text Box to PDF Page in C# Project with .NET PDF Library.
how to insert text in pdf reader; add text to pdf in acrobat
ChenandXie: OnlineConsumerReview:Word-of-MouthasaNewElementofMarketingCommunicationMix
ManagementScience54(3),pp.477–491,©2008INFORMS
479
expertusersisrelativelylargeandcostoftheproduct
islow.
Fromatheoreticalperspective,thispaperismostly
related to Lewis and Sappington (1994), which pro-
posesamodeltoshowwhenitisoptimalfortheseller
toprovidepartial-versusfull-attributeinformationto
consumers. In their model, there is only one infor-
mation channel between the seller and consumers
(i.e.,fromthesellertoconsumers).UnlikeLewisand
Sappington (1994), we allow anadditional informa-
tionchannel(i.e.,fromconsumerstoconsumers),and
examineaseller’sinformationdecisioninasettingof
dualchannels.
Substantially, this paper augments the traditional
marketing communications literature. To date, very
fewstudieshaveexaminedafirm’sstrategicdecisions
regardinginformationcontentforitsmarketingcom-
munications. Wernerfelt (1994b)and Simester (1995)
investigatewhenandhowfirmsshouldincludeprice
information in their advertising. Godes (2003)stud-
iestheimplicationsofthevalue-creatingversusper-
suasivepersonalsellingformat. ChenandXie(2005)
examine afirm’sadvertisingformat strategyin the
presenceofthird-partyproductreviews,andfindthat
using review-endorsed advertising (i.e., advertise-
mentscontainingthird-partyaward logos)to broad-
cast its success can hurt the winning product of a
productreview.Inthispaper,westudyafirm’sinfor-
mationcontent strategyby investigatinghow much
andwhattypeofproductinformationasellershould
providetoitscustomers.
The remainder of this paper is organized as fol-
lows:Section2illustratesthemarketingroleofonline
consumer reviews and how they differ from other
typesofproduct information;§3presentsourmodel
setup;§4examineshowthesellershouldbestrespond
to consumer reviews (i.e., the optimal information
contentdecision); §5studiesconditionsunderwhich
the seller should initiate or facilitate such consumer
reviewinformationitself(i.e., theoptimal consumer
reviewsupplydecision);and§6concludesthepaper
and discussessome strategicimplicationsand direc-
tionsforfutureresearch.
2. OnlineConsumer Review:
An EmergingSource of
ProductInformation
As an emerging source of product information,
what fundamental role can online e consumer review
play in the marketplace? How does online con-
sumer reviewdifferfromotherproduct information,
suchasseller-createdproductinformation,traditional
word-of-mouth (WOM), and third-party product
reviews?
2.1. OnlineConsumerReviewasaNewElement
intheMarketingCommunicationsMix
As consumer-created information, online consumer
review is likely to be more relevant to consumers
thanseller-createdinformation.Seller-createdproduct
information is more likely to be product oriented,
because it often describes product attributes in
termsoftechnicalspecificationsandmeasuresprod-
uct performance by technical standards. In con-
trast, the consumer-created product information is,
bydefinition, user oriented. It oftendescribesprod-
uct attributesintermsofusage situationsand mea-
suresproductperformancefromauser’sperspective
(Bickart and Schindler 2001). Consumers have dif-
ferentinformation-processingcapabilitiesininferring
benefitsfromproductattributeinformationduetodif-
ferentlevelsofexpertise(AlbaandHutchinson1987).
For this reason, seller-created product information
maybemoreusefultomoresophisticatedconsumers
(i.e.,experts).Consumer-createdproductinformation,
however, canhelp less-sophisticated consumers(i.e.,
novices)infindingtheirbest-matchedproducts.Asa
result, consumer reviews canbe e deployed asanew
element in the marketing communications mix and
canworkasanonline seller’sfree“salesassistants”
(Wernerfelt1994a)tohelpconsumerstoidentifyprod-
uctsthatbestmatchtheirneeds.
Consumer reviews are important for unsophisti-
catedconsumers(i.e.,novices), who mayhesitate to
purchaseifonlyseller-createdproductinformationis
available.However,thissalesassistantdoesnotcome
without cost. By allowing consumers to post their
own product evaluations, the seller creates a new
information channel for consumers, which thereby
eliminatestheseller’scapabilitytocontrolthesupply
ofproductinformation.Inthispaper,westudywhen
thesellershouldfacilitateconsumerreviewsandhow
itadjustsitsowncommunicationstrategyinresponse
toconsumerreviewinformation.
2.2. OnlineConsumerReviewvs.Traditional
(Offline)Word-of-Mouth
Onlineconsumerreviews,asconsumer-createdprod-
uct information, canbe viewed asa special type of
WOM(e.g.,GodesandMayzlin2004).Differentfrom
the traditional WOM, the influence ofwhichistyp-
icallylimited to a local social network (e.g., Brown
andReingen1987,Biyalogorskyetal.2001,Shi2003),
the impactofonline consumerreviewscanreachfar
beyond the local community, because consumers all
overthe world canaccessareview viathe Internet.
In addition, in general, traditional WOM is not a
direct decision variable e for the seller. However, the
recentdevelopmentofinformationtechnologyallows
asellertoeffectivelyinitiateandbroadcastconsumer
onlinereviewsviaitsownwebsite.Aseller canalso
C# PDF Annotate Library: Draw, edit PDF annotation, markups in C#.
installed. Support to add text, text box, text field and crop marks to PDF document. Able class. C#.NET: Add Text Box to PDF Document. Provide
how to enter text in pdf; how to enter text in pdf form
C# WPF PDF Viewer SDK to annotate PDF document in C#.NET
Text box. Click to add a text box to specific location on PDF page. Line color and fill can be set in properties. Copyright © <2000-2016> by <RasterEdge.com>.
add text box to pdf; add text boxes to a pdf
ChenandXie: OnlineConsumerReview:Word-of-MouthasaNewElementofMarketingCommunicationMix
480
ManagementScience54(3),pp.477–491,©2008INFORMS
license consumer reviews fromintermediaries (such
asEpinions.com),anddecidewhentoofferthemon
itswebsite(e.g.,c-source.com).Giventhewidespread
impact of consumer reviews, our paper investigates
how firms should adjust their marketing communi-
cation strategy to respond to this emerging source
ofWOMinformation.OurpapercomplementsWOM
literature byalsoexaminingthenewandpotentially
powerfulopportunityforaseller to helpconsumers
createanddisseminatetheirpersonalopinionsabout
the seller’s products. We look at the benefits and
drawbacksofencouragingor discouragingthisspe-
cial type of WOM information and provide insight
into firms’ decisions on when and how to provide
consumerreviews.
2.3. OnlineConsumerReviewvs.Third-Party
ProductReview
Another informationsourcecloselyrelated to online
consumerreviewisproductreviewfromthirdparties
(e.g.,CNET.com,caranddriver.com, PCMagazine, PC
World). As discussed in Chen and Xie e (2005), third-
party product reviews provide product information
usually based on lab testing or expert evaluations.
Third-partyproductreviewstendtofocusonproduct
attributeinformation(e.g.,performance,features,and
reliability)becausesuchinformationiseasiertoquan-
tifyandmeasure.Asaresult,third-partyreviewrat-
ingsarelikelytobecorrelatedwiththeperformance
oftheseattributes.Differentfromthird-partyreviews,
online consumer reviews are posted by usersbased
on their personal experiences, which can be highly
affected by their taste preferences as well as their
personalusage situations. Forthisreason, consumer
reviewsaremorelikelytofocusonwhetherandhow
aproduct matches aspecificindividual’spreference
andusagecondition.
To illustrate this difference, we conduct aprelim-
inary empirical study. We chose the digital camera
forourstudybecauseitisanidealproductcategory
to study online consumer and third-party product
reviews,forthefollowingreasons:(1)accordingtothe
ConsumerElectronicAssociation’sannualownership
study(Raymond2006),thedigitalcamerahasbecome
oneofthetopfivemostpopularconsumerelectronic
products;and(2)since2000theInternethasbeenthe
most popular channel for consumers to buydigital
cameras (Photo MarketingAssociation International
2001). We have collected the followingdatafor our
empiricalstudy:
(1) Third-partyproductreviewdatafromCNET.com,
the leading third-party professional review website
for consumer technologyproducts. Whenreviewing
digitalcameras,aCNET.comeditorpresentsdetailed
product attribute information, and rates the camera
onascaleof0through10based onhis/her evalua-
tionsonfourkeyaspectsofcameras:features,perfor-
mance,imagequality,anddesign.
(2) Consumer review data from Amazon.com, the
pioneerandtopproviderforonlineconsumerprod-
uct reviews. When posting reviews for a camera at
Amazon.com,consumersareaskedtogiveastarrat-
ing(from1 to 5) and write e a paragraph describing
theirexperiencesandrationalefortheirratings.Based
on the different consumer postings, Amazon.com
givesanaveragecustomerratingforeachmodel.
(3) Product attribute data fromCNET.com. We col-
lect data on the three most important digital cam-
eraattributessuggested byConsumer Reports: image
resolution (megapixels), optical zoom, and shooting
speed.
(4) Other controlvariables. Asthecontrolvariables,
wealsocollectdataontheproductlaunchdatefrom
CNET.comandthenumberofconsumerreviewpost-
ingsatAmazon.com.
Oursampleincludesall120digitalcameramodels
reviewedbyCNET.comfromJune2004toSeptember
2005.Table1(PartA)presentsthedescriptivestatis-
ticsofour samples.Asshowninthistable,for each
model in our sample, the average number of con-
sumerreviewspostedat Amazon.comis23,andthe
averageproductlengthoflife(thedifferencebetween
theproductlaunchdateandourdatacollectiondate)
is 338 days. Among 120 cameras, 90 models have
complete dataon third-party product review, prod-
uct attributes, and the two other control variables;
and 87 models have complete information on con-
sumerreviews,productattributes,andtheothercon-
trolvariables.
Using the CNET.com editor’s review ratings and
Amazon.comaverageconsumerratingsasthedepen-
dentvariables,weruntwoseparateregressionstosee
if the two types of product reviews—third-party
reviewandconsumerreview—haveasimilarrelation-
ship to product attribute information. As shown in
Table1(PartB),F statisticissignificantforthethird-
partyreviewmodelbutnotforthe consumerreview
model.Also,theratingofthird-partyreviewissignif-
icantlyaffectedbyopticalzoomandshootingspeed,1
butnoneofthethreeproductattributesaffecttherat-
ingoftheconsumerreview.Furthermore,forthe120
1The coefficient t of image e resolution isnot significant. One main
reason,assuggestedbyConsumerReports,isthattheimageresolu-
tion(megapixels)isthemajorcategoryvariableforadigitalcamera.
Therefore,reviewerstend torate differentcameraswithina cate-
gorysuchasfivemegapixelsinsteadofcomparingatwo-megapixel
model with a a five-megapixel model. . The positive e coefficient t for
product life e length showsthat,given the e same attribute e level, a
model can get higher ratingsif f it t was s launched d intothe market
earlier,which isconsistent with the fast-evolving characteristicof
thetechnology-drivendigitalcameramarket.
C# WinForms Viewer: Load, View, Convert, Annotate and Edit PDF
Add text to PDF document in preview. • Add text box to PDF file in preview. • Draw PDF markups. PDF Protection. • Sign PDF document with signature.
adding text to a pdf in acrobat; add text pdf reader
C# WPF Viewer: Load, View, Convert, Annotate and Edit PDF
Highlight PDF text in preview. • Add text to PDF document. • Insert text box to PDF file. • Draw markups to PDF document. PDF Protection.
how to add a text box to a pdf; how to insert text box in pdf
ChenandXie: OnlineConsumerReview:Word-of-MouthasaNewElementofMarketingCommunicationMix
ManagementScience54(3),pp.477–491,©2008INFORMS
481
Table1
Third-PartyProfessionalReviewvs.OnlineConsumerReview
A:Descriptivestatistics
N
Min.
Max.
Mean
Std.deviation
Third-partyreview
rating(CNET.com)
120
48
84
68
067
Consumerreview
rating(Amazon.com)
115
30
50
42
044
Imageresolution
(megapixels)
120
19
1
5 3
53
182
Opticalzoom(X)
111
00
1
0 2
41
262
Shootingspeed(fps)
98
00
50
21
092
Productlifelength
(days)
120
710
7390
3379
1
56
61
No.ofAmazon.com
consumerpostings
120
00
1
034
232
2397
B:Regressionanalysis
Dependentvariable:
Dependentvariable:
Third-partyreview
Consumerreview
rating(CNET.com)
rating(Amazon.com)
Standardized
Standardized
Independentvariables
coefficient tstatistic coefficient tstatistic
Imageresolution
0082
08310
112
0968
Opticalzoom
0356∗∗∗
3770
0088
0799
Shootingspeed
0264∗∗
2669
0017
0145
Productlifelength
0194
1910
−0089
−0756
No.ofAmazon.com
0010
0091
consumerreviews
No.ofobservations
90
87
R-squared
0247
0032
F-statistic
6986∗∗∗
0747
Note.Thespecificationalsoincludesanintercept.
p<010,∗∗p<005,∗∗∗p<001.
modelstested,thecorrelationbetweenthe consumer
reviewratingsandthird-partyproductreviewratings
isonly0.267(p<001),whichsuggeststhatthesetwo
typesofreviewsmaynotofferthesameinformation.
A closer look at the data finds that although
many camera models get low ratings from a third
party, they get high ratings from consumers. For
instance, Kodak Easy Share Z 740 gets 6.4 out 10
points rated byexperts at CNET.com, but gains 4.5
out 5stars based onmore e than 100consumer post-
ingsatAmazon.com.Whereastheprofessionalreview
focusesonitslukewarmperformance,mostconsumer
reviews praise how this camera matches their dif-
fering usage e conditions. For instance, in the e Kodak
EasyShareZ740example,justbrieflyskimming“the
most recent 20 postings” at Amazon.com, we can
identifymorethan10differentusagesituations,vary-
ing from outdoor landscape shots, animal and bird
shots,kids’sports,NewYear’sEvecelebrations,long-
distanceshooting,Christmasgifts, overseasvacation
trips, sharing photos with friends, family trips to
DisneyWorld, even crime-scene e photography. Here
are wordsinsometypicalconsumerreviews:“Iwas
looking for something that took the picture NOW!
As opposed to 3 seconds later. Beinga parent, this
was very important to methis camera is a real
treasure”;“The10Xzoommakesiteasytoseeimages
a long way away. I am able to capture the beauty
of deer and other outside landscapes and animals
at great quality”Differently, focusingmainlyon
the attribute information,atypicalparagraphinthe
third-partyreviewfromCNETforthesamemodelis
“shutterlagwasmoderateat0.7secondunderhigh-
contrastlightingbutalanguorous1.8secondsunder
morechallenginglow-contrastlighting,evenwithaid
fromthefocus-assistlamp.”Basedonthisqualitative
inquiry, it isclearthat consumersevaluatetheirpur-
chased product based on whether it fits their indi-
vidualpreferencesandperformswellintheirspecific
situations, which is quite different from third-party
reviewsprovidedbyprofessionalsemphasizingprod-
ucttechnicalspecificationsandperformances.
3. ModelSetup
Inthissectionwespecifykeyassumptionsandsetup
forourmodel. Keynotationsaresummarizedinthe
appendix.
3.1. SellerandConsumers
We consider a monopoly seller2 carrying a multi-
attribute product. Let c denote e the marginal cost of
theproduct.
We allow consumer heterogeneity in two dimen-
sions: preference and expertise. First, we allow con-
sumers to differ in their preferences toward the
seller’sproduct.Foragivenproduct,someconsumers
will find that the product matches their preference
better than others. Specifically, consider a product
with two attributes, a
1
and a
2
. For any given con-
sumer,thereisanequalchancethatagivenattribute
matches her preference, which is known to both
the sellerand buyers.Consumerpreferencesfor the
twoattributesareindependent.Forexample,avideo
game oftenhas two keyattributes: (1)genre, which
specifiesthetypeofthegame(i.e.,role-playinggame
or strategy game); and (2) plot difficulty, which
indicatesthechallengeleveltothe players.Indepen-
dence inpreference impliesthat a consumer’s pref-
erence for the type of the game is not necessarily
relatedtoherpreferenceforthedifficultyofthegame
2The seller’smonopoly position mainly results s from
consumers’
loyalty and limited search. Recent studies have demonstrated
onlineconsumers’loyaltyandlimitedsearchforonlinesellers.For
example,Johnsonetal.(2004)presentempiricalevidencethatcon-
sumeronlinesearchisverylimitedduringthe shopping process.
On average,consumersvisit 1.2 booksitesand 1.3 compact disc
sitesineachcategory.Themonopolymodelcanhelpusunderstand
thefundamentalimpactofthenewinformationchannel,i.e.,online
consumerreviewonfirmmarketingstrategies(e.g.,Shugan2002).
C# PDF Sticky Note Library: add, delete, update PDF note in C#.net
Allow users to add comments online in ASPX webpage. Able to change font size in PDF comment box. Able to save and print sticky notes in PDF file.
how to insert text box on pdf; add text to a pdf document
.NET PDF Document Viewing, Annotation, Conversion & Processing
PDF Write. Insert text, text box into PDF. Edit, delete text from PDF. Insert images into PDF. Edit, remove images from PDF. Add, edit, delete links. Form Process
adding text fields to pdf acrobat; adding text to pdf
ChenandXie: OnlineConsumerReview:Word-of-MouthasaNewElementofMarketingCommunicationMix
482
ManagementScience54(3),pp.477–491,©2008INFORMS
(i.e., a role-playing game lover may prefer difficult
or easy games). Hence, consumers can be catego-
rized into four types according to their preference-
matchingsituationswith theproduct: fullymatched
type T
MM
(matching on both attributes), partially
matched typesT
M
M
and T
MM
(matchingonattribute
a
1
ora
2
),andfullyunmatchedtypeT
M
M
(matchingon
neitherattribute).LetvFvPvdenoteconsumerval-
uationsforfullymatched,partiallymatched,andfully
unmatched consumers when consumers have com-
plete informationonproductattributes,respectively.
Whenconsumersarenot fullyaware ofinformation
onproductattributes,theyformtheirvaluationbased
ontheexpectedvalue.Forexample,inthecasewith-
outanyproductinformation,allconsumershavethe
samewillingnesstopay,v=v+2vP+v0/4.Without
lossofgenerality,weassumev0=0.
Second,weallowconsumerstodifferintheirexper-
tise and knowledge about the product. We consider
two consumersegments: an expert segment (E)and
a novice segment (N). Let S denote the segment,
where S=EN. Albaand Hutchinson (1987,p. 428,
Proposition 4.7) argue that, due to their difference
incausalinference capability,“expertsaremoreable
thannovicestoinferintended productbenefitsfrom
technical information and to infer likely technical
causesofclaimedbenefits.”Thispropositionimplies
that experts are more likely than novices to cor-
rectly map their usage situations with the product
attributesbased onthe attribute informationoffered
bythe seller. Because consumer-created information
is more user oriented than seller-created informa-
tion, it islikelythat all consumerscanbenefit from
suchinformation. Accordingly,we assume that,sim-
ilar to experts, novices can identify a matching or
mismatching product by learning from the experi-
ence of existing users through consumer reviews.
However, different from experts who can find the
matchedproductsolelybasedonseller-createdinfor-
mation,noviceconsumersareunabletomatchprod-
uctattributeswiththeirpreferencesintheabsenceof
consumerreviews.
Todistinguishconsumerheterogeneityinthepref-
erence and expertise dimensions discussed above,
hereafterwerefertoconsumerswithadifferentpref-
erence as a different consumer type and consumers
withadifferentexpertiselevelasadifferentconsumer
segment. Asdescribedearlier,thereare four typesof
consumers with different preference-matching situ-
ations (i.e., T =T
MM
T
M
M
T
MM
T
M
M
) and two seg-
ments of consumers with different expertise levels
(i.e., S=EN). The preference dimension and the
expertise dimension are orthogonal, i.e., for both
expertandnovicesegments,therearefourtypesofcon-
sumerswithdifferentpreference-matchingsituations.
3.2. InformationStructure
We allow a two-sided information asymmetry be-
tweenthe seller and consumers. The e seller has pri-
vateproductinformation,buthasnoinformationon
consumercharacteristics.Consumersknowtheirown
tastesand expertise levels, but have no information
onproductattributes.
Therearetwopossibleinformationsourcesforcon-
sumers: (1) seller-created product attribute informa-
tion, (2)consumer review information(if available).
Due to the seller’s concern for its reputation, we
assume the seller-created product attribute informa-
tionis accurate e and truthful. Without loss ofgener-
ality,we assume theseller’sinformationsupplycost
iszero,consideringthesignificantlyreducedcostsof
collectingand distributinginformationviathe Inter-
net(Averyetal.1999).
Theinformationstructureisdeterminedby(1)how
muchattributeinformationthesellerprovidesviaits
owncommunicationto consumers, and (2)whether
productreviewscreatedbythecurrentusersareavail-
able to consumers. We call the former information
content strategy and the latter consumer review sup-
ply strategy. The information content strategy is com-
pletelydeterminedbytheseller.Specifically,theseller
can choose to adopt a full-information strategy (i.e.,
providinginformation on both attributes), a partial-
informationstrategy(i.e., providinginformationonly
on one of the two attributes), or a no-information
strategy. To focus on more realistic and interesting
cases (i.e., full- and partial-information strategy), we
assumethatintheabsenceofconsumerreviewinfor-
mation,theno-informationstrategycannotbeoptimal.
Specifically, we assume e that the seller cannot make
a positive profit when refusing to offer any prod-
uctattributeinformationtoconsumers.LetI=IFIP
denotetheinformationcontentdecision,whereIF and
IP present the case when the e seller adoptsfull- and
partial-informationcontentstrategies,respectively.
Let P denote e the probabilitythat aconsumer’s
true status is  for an attribute, where =M
M
(i.e.,theattributecanbeeitheramatchormismatch).
Intheabsenceofanyinformationabouttheattribute,
thereisanequalchance (i.e., 50%)thatthe attribute
isamatch/mismatchforaconsumer.Hence,PM=
P
M=1/2.
Let  denote the informativeness (accuracy and
content) of consumer review information, where
0≤≤1.Ahigherdegreeofinformativenessisasso-
ciated with better information such that the con-
sumer review information is perfectly informative
when =1, and completely uninformative when
=0. Let Ps denote e the conditionalprobability
3Thisassumption impliesthat t v<
c,where v v isbuyer expected
valueintheabsenceofanyproductinformation.
VB.NET PDF - Annotate PDF with WPF PDF Viewer for VB.NET
Text box. Click to add a text box to specific location on PDF page. Line color and fill can be set in properties. Copyright © <2000-2016> by <RasterEdge.com>.
adding text to a pdf; how to add text to a pdf file in acrobat
VB.NET PDF - Annotate PDF Online with VB.NET HTML5 PDF Viewer
Click to add a text box to specific location on PDF page. Outline width, outline color, fill color and transparency are all can be altered in properties.
adding text to pdf form; how to insert text in pdf reader
ChenandXie: OnlineConsumerReview:Word-of-MouthasaNewElementofMarketingCommunicationMix
ManagementScience54(3),pp.477–491,©2008INFORMS
483
thatabuyer’struevaluationisafterreceivingasig-
nal s from consumer reviews, where e s=mm(e.g.,
the signal can be e either a “match” or “mismatch”).
Intuitively,Psdependsonthedegreeofinforma-
tivenessofconsumerreviewssuchthatconsumersare
morelikelytocorrectlyidentifytheirtruestatus(i.e.,
matchormismatchwiththegivenattribute)basedon
consumerreviewswhenthereviewsismoreinforma-
tive.Formally,
dPMm/d>0 dP
Mm/d>0
dP
Mm/d<0 dPMm/d<0
(1)
Formathematicaltractability,weadoptthesamespec-
ificationofPssuggestedinLewisandSappington
(1994):
PMm=P
Mm=q¯=1/2+/2
PMm=P
Mm=q
=1/2−/2
(2)
Thisspecificationsatisfies(1)andimplies:
(a) When reviews are completely uninformative
(=0),consumers’truestatusisindependentofthe
signal received from consumer reviews, and such
reviewsareuseless(i.e.,PMm=PM=1/2.
(b) When reviews are completely informative
=1, consumers can perfectly identify their true
status based on consumer review information (i.e.,
PMm=P
Mm=1.
(c) When
reviews are partially
informative
0<<1,consumersbenefitfromconsumerreview
informationbutareunabletoperfectlyidentifytheir
true status based on it (i.e., PM<PMm<1
P
M< P
M  m <10 < < P
M m< < P
M, and
0<PM
m<PM.4
Note that (a)–(c)implythat the buyer revisesher
expectedvaluationupwardafterreceivinga“match”
signal, but downward after receiving a “mismatch”
signal.Themoreinformativethereviewinformation,
themorethebuyeradjustshervaluationbasedonthe
consumerreviews.
It is important to notice that although the seller
hasfullcontrolovertheinformationcontentdecision,
theavailabilityofconsumerreviewsmaynotbecom-
pletelytheseller’sdecision.Forexample,evenifthe
seller determinesnot to offer its usersthe e optionto
4Asshown in Appendix x A.2, the e specification in
(2) can n be de-
rived asa posterior probabilitybyspecifyingPsPmM=
Pm
M=1/2+/2 and PmM=Pm
M=1/2−/2,
wherePsistheconditionalprobabilitythata buyerobtainsa
signalsfromconsumerreviewsgiventhatabuyer’struestatusis.
Such aspecification ofPsimpliesthat(a)theprobabilitythat
thebuyergetsthecorrectsignalfromconsumer reviewsincreases
withandapproachesto1when=1,and(b)theprobabilitythat
thebuyergetsanincorrectsignalfromconsumerreviewsdecreases
withandapproachesto0when=1.
posttheirproductreviewsonitswebsite,somethird
parties(e.g.,consumersocialnetworkorinfomediary
websites) may decide to create such user-generated
information for a given product at any time. As a
result,somesellersmayfindthemselvesfacingunex-
pected consumer reviews (i.e., the case e of the e seller
asanobserver ofonline WOMinGodeset al. 2005).
In §4.2, we examine this case and show how the
sellerinthissituationcanbestrespondtounexpected
consumer reviews byadjustingits own information
content strategyonce reviews become e available. We
call this defensive response to consumer reviewers.
In§4.3, weexamine the casewherethe sellerantici-
patestheavailabilityofconsumerreviews(e.g.,ifthe
seller allows consumersto post their reviews onits
ownwebsite),andderivetheseller’sproactivestrategy
towardconsumerreviews.
3.3. ModelTiming
We consider two periods, t=12. In each period,
oneunitofconsumers(withdifferentpreferencesand
expertise levels)arrivesatthe market, makesapur-
chase decision,andthenexits. Let  denote the size
ofthe experts over two periods, and 
t
denote the
fractionofexpertsinperiodt(i.e.,1−
t
isthe frac-
tionofnoviceconsumers),where
t
=
t
,1≥
1
>0,
1≥
2
≥0and
2
t=1
t
=1.Period1arrivalscanlearn
aboutproduct attributesonlyfromthe seller-created
information. Period 2arrivalscanlearnaboutprod-
uct attributesfromanadditionalinformationsource,
consumerreviewinformation,ifsuchreviewsbecome
available in period 2. In each period t, the seller
adjustsitsinformationcontentstrategyI andpriceP
t
to determine which differentsegmentsandtypes of
consumerstoserve.
4. InformationResponseto
Consumer Reviews
Inthissection,weexaminetheseller’sbestresponse
toconsumerreviews.Wefirstderivetheseller’sopti-
mal strategy in the absence e of consumer review in
§4.1. We then study the seller’s defensive response
andproactiveresponsetoconsumerreviewsin§§4.2
and4.3,respectively.
4.1. Benchmark:IntheAbsenceof
ConsumerReviews
We first derive conditions under which it is opti-
malforthesellertosupplypartial(full)attributein-
formationto consumersinthe absence ofconsumer
reviews.
5We assume
1
>0 (a nonzero number of f experts arriving in
period1)toensuretheavailabilityofconsumerpostings.
ChenandXie: OnlineConsumerReview:Word-of-MouthasaNewElementofMarketingCommunicationMix
484
ManagementScience54(3),pp.477–491,©2008INFORMS
Partial-InformationStrategy. Underthisstrategy,the
sellerprovidesinformationononlyoneattribute,for
instancea
1
. Expertsare certainabouttheirmatchor
mismatchonthe informedattribute (a
1
), but remain
uncertainabout the uninformedattribute(a
2
). With-
outanyinformationontheuninformedattribute,they
perceive an equal probability that the uninformed
attributeisamatchormismatch.Hence,thevaluation
is (vP +vF/2 for experts whose taste matches the
informedattribute(i.e.,T
M
M
andT
MM
)andvP+v0/2
for experts whose taste mismatches the informed
attribute(i.e.,T
MM
andT
M
M
),respectively.Duetotheir
inabilitytoprocesstheseller-createdinformation,all
noviceconsumersperceiveanequalprobabilitytobe
one of four possible types and their expected valu-
ation is
v=v+2v+v0/4. The seller maximizes
itsprofitsbysettingitsprice.Inordertodifferentiate
thebenchmarkcasefromthatinthepresenceofcon-
sumerreviewstobeexaminedbelow,weusea“hat”
forallthe variablesintheformer.The optimalprofit
underthepartial-informationstrategyinthe absence
ofconsumerreviewsis
Ip.
Full-Information Strategy. Under this strategy, the
seller provides information on both product at-
tributes. Expert consumersarefullyinformed. Their
valuationsisvF forfullymatched(T
MM
),vP forpar-
tially matched (T
M
M
, T
MM
), and v0 for mismatched
(T
M
M
) experts, respectively. The expected valuation
for novicesis
v. The optimalprofit under full-infor-
mationstrategyin the e absence ofconsumer reviews
is
IF.
Let
ˆ
I denote the optimal information content
strategyintheabsenceofconsumerreviews.Compar-
ingtheprofitsunderthetwostrategies,partialinfor-
mationand full information, leadstoLemma 1(see
proofsofalllemmasandpropositionsintheappendix
andtheonlinetechnicalappendix,whichisprovided
inthee-companion).6
Lemma1(Benchmark). In the absence of consumer
reviews,theseller’soptimalinformationcontentandpric-
ingstrategyis
offeringfullinformation
i.e.
ˆ
I=IFP
t
=vF
ifc≥vP
offeringpartialinformation
i.e.
ˆ
I=IPP
t
=v+vF/2 otherwise
Lemma1revealsthat,inthe absence ofconsumer
reviews, either afull- or a partial-information strat-
egy can be e optimal. Note that full information is a
6An
electronic companion to this paper is s available as part of
the online version that can be found at http://mansci.journal.
informs.org/.
margin-driven strategy, because by offering informa-
tiononbothattributes, the seller isable tocharge a
high price, vF, to the fullyinformed/fullymatched
experts, although other consumers will be priced
out of the market at this high price. On the other
hand, partial information is a volume-driven strat-
egy because when the seller offers information on
onlyoneattribute,bothfullymatchedandsomepar-
tiallymatchedexperts(i.e.,expertswhosepreference
matchestheinformedattributea
1
)perceivethesame
probabilityto be the e fullymatched type. The seller
can sell to bothsegmentsbycharginga sufficiently
low price, v+vF/2, to compensate for the con-
sumers’uncertainty. Lemma1showsthat it isopti-
maltoofferfullinformationandonlyservethefully
matched experts (i.e., pursue e a margin-driven strat-
egy)when the e cost is sufficientlyhigh (c≥vP), but
offer partialinformation and serve more e types (i.e.,
pursueavolume-drivenstrategy)otherwise.
Lemma1isconsistentwithLewisandSappington
(1994), although neither the buyer’s heterogeneity
in expertise e nor the availabilityofconsumer review
informationisconsideredintheirmodel.Intherestof
thissection,weexaminehowthesellershouldadjust
itsinformationcontentstrategygiveninLemma1in
responsetoconsumerreviews.Then,inthenextsec-
tion,we examine theseller’sdecisiononthesupply
ofconsumerreviews—bothissuesthathavenotbeen
previouslyexplored.
4.2. TheOptimalDefensiveInformation
ResponsetoConsumerReviews
Wenowconsiderthecasewherethesellerfacesunex-
pected reviews, and show how the seller can opti-
mally adjust its own information content strategies
inresponse to consumer reviews.Inthe presence of
consumer reviews, period 1arrivals learn about the
product only from seller-createdinformation, hence,
havingthesamevaluationasinthebenchmarkcase.
Accordingly,theseller’speriod1strategyisthesame
as that given inLemma 1. However, different from
the benchmarkcase, period2arrivalscannowlearn
about the product not onlyfromthe seller, but also
fromexistingbuyersviaconsumerreviews.Theirval-
uations in period 2 may be different from that in
period1,asshownbelow.
SupplyingPartialAttributeInformationI=IP
Expert Consumers. In period 2, consumer reviews
with informativeness  become available. Hence,
period 2 experts can use the signal from con-
sumer reviews to update their matching probabil-
ity on the uninformed attribute a
2
. For example,
for those experts who have e a match with informed
attribute a
1
 and receive e the match signal onunin-
formed attribute a
2
, their expected valuation is
q¯v+q
vP.Similarly,wecanderivetheexpected
ChenandXie: OnlineConsumerReview:Word-of-MouthasaNewElementofMarketingCommunicationMix
ManagementScience54(3),pp.477–491,©2008INFORMS
485
valuations for other types of period 2 experts (see
AppendixA.2).
Novice Consumers. In period 2, novice consumers
updatetheirvaluationbasedonconsumerreviewin-
formation. For example, with the help of consumer
reviewinformation,thenoviceswhoreceivethematch
signalsonbothattributeshaveastheirexpectedvalu-
ationq
q
v0+2
¯
qq
vP+
¯
q
¯
qvF.Similarly,
wecanfindtheexpectedvaluationsforothertypesof
novices(seeAppendixA.2).
Giventhevaluationsofdifferent consumer types/
segments,thesellersetspriceP
2
tomaximizeitsprofit
inperiod2,
2
IP.
SupplyingFullAttributeInformationI=IF
Expert Consumers. If the seller adopts the full-in-
formation strategy, experts in period 2 are fullyin-
formed. Their valuation is vFvPv0 for types T
MM
,
T
MM
andT
M
M
,andT
M
M
,respectively.
NoviceConsumers. Althoughthesellerprovidesfull
information, the expected valuations of novices are
thesameasinthecasewherethesellerprovidespar-
tialinformation,providedthatnoviceconsumersare
unabletoprocessseller-createdinformation.
Giventhevaluationsofdifferent consumer types/
segments,thesellersetspriceP
2
tomaximizeitsprofit
inperiod2,
2
IF.7
We now examine WHEN and HOW the seller
should varyitsowninformationcontent strategyin
responsetoconsumerreviews.LetI∗ denotetheopti-
malinformationcontentstrategiesinthepresenceof
consumer reviews and I denote the difference in
theoptimalamountofinformationcontentwithand
without consumer reviews, I=I∗ 
ˆ
I. The follow-
ingpropositionstatestheseller’soptimalinformation
responsetoconsumerreviews.
Proposition1(DefensiveInformationResponse
to Consumer Reviews). Facing unexpected consumer
reviews, the sellercanimproveits profitbyadjustingits
own information content strategy once reviews become
available.Specifically,comparedwiththecasewithoutcon-
sumerreviews,inthepresenceofconsumerreviews,itis
optimalforthesellerto
(a) INCREASEattributeinformationviaitsowncom-
munication if two conditions hold: (i) the product cost
is low, and (ii)either the review informativeness is in a
midrange,orthereviewinformativenessisextremelyhigh
andtherearesufficientexpertconsumers.
(b) DECREASEattributeinformationviaitsowncom-
municationifthreeconditionshold:(i)theproductcostis
high,(ii)thereviewinformativenessisinamidrange,and
(iii)therearesufficientnoviceconsumers.
7Theoretically, the seller can
also decide to supply no o attribute
information in the presence ofconsumer reviews.However, , it t is
straightforwardthatthisstrategyisadominatedstrategycompared
withthepartial-informationstrategy.
(c) MAINTAINthesamelevelofattributeinformation,
otherwise.
MathematicallyI=I
ˆ
I
=
I
F
−I
P
>0 ific<v
P
(ii)∈

or
∈
1and>
N
IP−I<0
ific≥v
P
ii∈


iii<
N
0 otherwise
(3)
Equation(3)impliesthatthesellercanbestrespond
to the availability of consumer reviews by revers-
ing its information content strategy, i.e., switching
from supplying full to partial information or vice
versa. This result is intriguing because it implies
that the two types of information, consumer cre-
ated and seller created, can be either complements
(i.e., consumer reviews increase the seller’s incen-
tive to supply attribute information) or substitutes
(i.e., consumer reviews decrease the seller’s incen-
tive to supply attribute information). Proposition 1
reveals that the existence and direction ofthe inter-
action are determined bythree e product/market fac-
tors:(1)productcost,(2)informativenessofconsumer
reviews,and(3)thesizeofdifferentsegmentsofcon-
sumers(expertsornovices).
For low-cost products, it is optimal to offer par-
tial information and sell to both fully and some
partially matched experts at a lower price in the
absence ofconsumer reviews (see e Lemma 1). How-
ever, such a volume-driven strategy is no longer
optimal when consumerscanlearn about the prod-
uct from consumer reviews with sufficient informa-
tiveness ∈

. This is because such review
information significantly decreases the valuation of
theexpertsreceivingthemismatchsignalontheunin-
formedattribute,evenasit alsoincreasesthevalua-
tionofthoseexpertsreceivingthematchsignalonthe
uninformed attribute. As a result, to maintain such
avolume-drivenstrategy,the sellerhasto reduceits
pricegreatly.Incontrast,switchingtoamargin-driven
(i.e.,full-information)strategy,underwhichtheseller
sells only to the fully matched experts at a high
price, is more profitable. Note that when consumer
reviewinformationishighlyinformative∈
1,8
the valuation ofnovices who receive match signals
on both attributes increases significantly, so that it
becomesprofitableto selltothese novice consumers
iftheirsegmentissufficientlylarge.Ifthenoviceseg-
mentissufficientlysmall(or equivalently,theexpert
8When=1,theseller isindifferent betweenprovidingfull,par-
tial,ornoinformationgivenazerosupplycost.
ChenandXie: OnlineConsumerReview:Word-of-MouthasaNewElementofMarketingCommunicationMix
486
ManagementScience54(3),pp.477–491,©2008INFORMS
segmentissufficientlylarge,>
N),switchingtothe
full-information strategy to only serve full-matched
expertsismore attractive e than retainingthe partial-
informationstrategy.
Forhigh-costproducts,intheabsenceofconsumer
reviews,amargin-drivenstrategy(offeringfullinfor-
mationand sellingonlyto fullymatched experts at
ahighprice)ismoreprofitablethanavolume-driven
strategy(offeringpartialinformationandsellingboth
tofullyandsomepartiallymatchedexpertsatalow
price) (see Lemma 1). However, in the presence of
consumer reviews, the volume-driven strategy can
be more attractive than the margin-driven strategy.
When consumer review information is sufficiently
informative≥
,thevaluationofthenoviceswho
receivematchsignalsonbothattributesincreasessuf-
ficiently,sothatitbecomesprofitableforthesellerto
reduce its price e to sell to these consumersin addi-
tiontotheperfectlymatchedexpertsifthenumberof
novicesissufficientlyhigh(equivalently,thenumber
ofexpertsissufficientlysmall,<
N).Asasidebene-
fitofthislow-price,volume-drivenstrategy,theseller
can profit from switching to the e partial-information
strategybygainingextrademandfromsomepartially
matchedexperts.
Notethatswitchingtothepartial-informationstrat-
egy will not help the seller if the e consumer review
information is extremelyinformative >). This
isbecause,underthepartial-informationstrategy,the
valuationofexpert consumerswho receive the mis-
match signal on the uninformed attribute is nega-
tivelyrelatedtothe reviewinformativeness(i.e.,the
more informative the reviews, the more likely for
these consumers to realize their mismatch). As a
result,whenthereviewinformativenessisextremely
high>,thevaluationoftheseexpertsbecomes
too low to retain the advantage of the partial-
informationstrategy.
4.3. ProactiveResponsetoOnline
ConsumerReviews
Wenowexaminethecasewherethesellercanproac-
tivelyrespondtoconsumerreviews.Thisispossible
when the seller anticipates the availability of con-
sumer reviewers in the second period (e.g., if the
seller allowsconsumersto post their reviews on its
ownwebsite). Different fromthe last section, where
the review informativeness  is exogenouslygiven,
the seller may now influence  by controlling the
number of consumers who purchase in period 1
(e.g., D
1
).9 This isbecause the e number ofbuyers in
9An
alternative way y to model a firm’s dynamic c behavior is to
considerconsumers’strategicwaitingbehaviorexplicitlyandtreat
t
asendogenous.Similartosome previousresearch(e.g.,Lazear
1986,Rajuetal.1990),ourmodeldoesnotexplicitlyconsidersuch
period1determinesthenumberofpotentialreview-
ers in period 2, which can be positively related to
theinformativenessofconsumerreviewinformation,
i.e., /D
1
>0.10 Note e that the sellercancontrolD
1
viaitsperiod1informationcontentstrategy(i.e.,the
margin-drivenstrategycharacterized byofferingfull-
informationandhighpricesorthevolume-drivenstrat-
egycharacterizedbyofferingpartial-informationand
lowprices).Let=D
1
=
1
,whichisthe levelof
reviewinformativenessreachedunderthemaximum
period1demand(i.e.,whenallperiod1arrivalsbuy
inperiod1under positive prices). We refer to the
reviewinformativenesspotential.
Whenthe selleranticipatestheavailabilityofcon-
sumer reviews inperiod 2, the seller canact proac-
tively to adjust its strategies even before reviews
become available, whichcan affect review informa-
tiveness.Proposition2stateswhenandhowtheseller
canbenefitfromthisproactivestrategy.
Proposition 2 (Proactive Strategy Towards
Consumer Reviews). When the seller anticipates the
availabilityofconsumerreviewers,itisoptimalto
(a) Adjustitsperiod1strategy(i.e.,pricesandinforma-
tioncontentsupply)ifthereviewinformativenesspotential
is sufficientlyhigh. Specifically, iff>
,thencompared
withLemma1,itisoptimaltochangeperiod1strategyby
decreasingprice
ifc<vP
switchingtopatial-informationstrategy
orremainingwithfull-information
strategybutreduceprice
ifc≥vP
(b) Adoptthesameperiod2responsegiveninProposi-
tion1.
Whentheselleradoptsaproactiveresponsestrategy
and adjusts its marketingstrategy before e consumer
reviewsbecome available, itwill offera lower price
andgeneratesahigherdemandinperiod1.Although
sucha response e reducesperiod 1profits, it cansig-
nificantlyincrease period 2 profits because a larger
numberofbuyersinperiod1leadsto ahigherlevel
ofinformativenessintheconsumerreviewsinperiod
2. Proposition4revealsthat whetheraseller should
adoptsuchproactivestrategydependsonthereview
(WOM)informativenesspotential,.Whenthereview
behavior.Animplicitassumptionbehindthistypeofmodelisthat
thediscountrateofearlyarrivalsisveryhigh,whichisconsistent
withthebehavior ofearlyconsumersinmanynewproduct mar-
kets(Moore1991).Ifthediscountrateofearlyarrivalsisverylow
andreacheszero(
t
isendogenous),someperiod1expertsmight
waittobuyinperiod2.Inthiscase,therequiredpricetoinduce
period1purchasewouldbelower.
10Wedonotspecificallymodelconsumerreview
postingbehavior
here.AdmatiandPfleiderer(2004)havestudiedconsumerposting
behaviorandreviewinformativeness.
Documents you may be interested
Documents you may be interested