The Field behind the Screen: Using Netnography for Marketing Research in Online Communities Author(s): Robert V. Kozinets Source: Journal of Marketing Research, Vol. 39, No. 1 (Feb., 2002), pp. 61-72 Published by: American Marketing Association Stable URL: http://www.jstor.org/stable/1558584 . Accessed: 30/09/2014 11:21
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

.

American Marketing Association is collaborating with JSTOR to digitize, preserve and extend access to Journal of Marketing Research.

http://www.jstor.org

This content downloaded from 137.43.71.99 on Tue, 30 Sep 2014 11:21:55 AM All use subject to JSTOR Terms and Conditions

ROBERTV. KOZINETS*
The authordevelops"netnography" an onlinemarketing as research consumerinsight.Netnography ethnography is techniquefor providing to As is adapted the studyof onlinecommunities. a method, netnography and andmore faster,simpler, less expensivethantraditional ethnography naturalistic unobtrusive focus groupsor interviews. provides and than It information the symbolism, on of meanings,and consumption patterns onlineconsumergroups.The authorprovidesguidelines that acknowland edge the onlineenvironment, respectthe inherent flexibility openness of ethnography, provide and and ethicsin the conductof marrigor keting research.As an illustrative example, the author providesa of and netnography an onlinecoffee newsgroup discusses its marketing implications.

The in

Field Online

Behind

the

Screen:

Using
Research

for Netnography Marketing Communities

Consumersmakingproductand brandchoices are increasingly turning to computer-mediatedcommunication for informationon which to base theirdecisions.I Besides perusing advertisingand corporateWeb sites, consumersare using newsgroups,chat rooms, e-mail list servers, personalWorld Wide Web pages, and other online formats to share ideas, build communities, and contact fellow consumers who are seen as more objective informationsources. Although they are popularly called "virtual communities" (Rheingold 1993), the term"virtual" might misleadinglyimply thatthese communities are less "real" than physical communities (Jones 1995). Yet as Kozinets (1998, p. 366) points out, "these social groups have a 'real' existence for their participants, and thus have consequentialeffects on many aspects of behavior,including consumer behavior"(see also Muniz
IFor example, surveys of adults who use online services indicate that 36% of them access newsgroupsand 25% visit chat rooms (Visgaitis 1996), and these numbersappearto be growing (Jones 1999). Reid's (1995) analysis of Arbitrondata provides a much higher figure: 71.6% of all Internet users assess newsgroups. *RobertV. Kozinets is Assistant Professorof Marketing,Kellogg School of Management, Northwestern University (e-mail: r-kozinets@kellogg. nwu.edu; Web page: http://www.kellogg.nwu.edu/faculty/kozinets/htm/n research). The author thanks the marketing seminar group at Kellogg, Annama Joy, Jay Handelman,and John Sherry for comments on previous versions of this article.The threeJMRreviewersalso providedkindencouragement and useful remarksthat helped improve the article. The members of the alt.coffee newsgroup generously contributedtheir utterancesand insights.

and O'Guinn 2001). To maintainthe useful distinction of social gathering,I use the term "online computer-mediated communities"to refer to these Internet-based forums. Motion pictures, sports, music, automobiles, fast food, softtoys, consumerelectronics, computersand peripherals, ware,cigars, beer,coffee, and many otherproductsand services are discussed in online communitieswhose importance is being increasinglyrecognizedby contemporary marketers (see, e.g., Armstrong and Hagel 1996; Bulik 2000; Hagel and Armstrong 1997; Kozinets 1999; Muniz and O'Guinn 2001; White 1999). In the past few years, marketingfirms such as Cyveillance, eWatch, NetCurrents,and GenuOne and consumer services such as Epinions.com, PlanetFeedback, Bizrate.com,and eComplaints.comhave been formed to take advantageof opportunitiesposed by cross-consumer electronic communication. The reason behind this marketing interest is twofold. First, marketersrecognize the increasingimportanceof the Internetand of consumerswho are active in online communities. Almquist and Roberts (2000, p. 18) find that the majorfactor influencingpositive brandequity for one brand over anotheris consumeradvocacy.Online communitiesare contexts in which consumers often partakein discussions whose goals include attemptsto informand influence fellow consumers about products and brands (Kozinets 1999; Muniz and O'Guinn 2001). Second, one of the major purthe poses of marketingresearchis to identifyand understand tastes, desires, relevant symbol systems, and decisionmaking influences of particularconsumers and consumer groups. As the advent of networkedcomputing is opening
Journal of MarketingResearch Vol. XXXIX (February2002), 61-72

61

This content downloaded from 137.43.71.99 on Tue, 30 Sep 2014 11:21:55 AM All use subject to JSTOR Terms and Conditions

62

OF JOURNAL MARKETING 2002 RESEARCH,FEBRUARY researcherswith a window into naturallyoccurringbehaviors, such as searches for information by and communal word of mouth between consumers.Because it is both naturalisticand unobtrusive-a uniquecombinationnot found in any other marketingresearchmethod-netnography allows online social continuingaccess to informantsin a particular situation.This access may provide importantopportunities for consumer-researcherand consumer-marketerrelationships. The limitations of netnographydraw from its more narrowfocus on online communities,the need for researcher skill, and the lack of informantidentifierspresinterpretive ent in the online context that leads to difficulty generalizing results to groups outside the online community sample. Marketingresearcherswishing to generalize the findings of a netnographyof a particularonline group to other groups must therefore apply careful evaluations of similarity and employ multiple methods for triangulation. In this article'sfirst section, the methodof netnographyis explained, and particularattention is paid to its relative strengthsand weaknesses comparedwith in-personqualitative techniques.The second section provides an illustrative example that uses the informationon a popularcoffee newsgroup to gatherconsumer insights that may inform marketing practice. THEMETHODOF NETNOGRAPHY and Netnography Ethnography Ethnography is an anthropological method that has gained popularityin sociology, cultural studies, consumer research,and variousother social scientific fields. The term refersboth to fieldwork,or the study of the distinctivemeansocial groups, and ings, practices,and artifactsof particular to the representations based on such a study.Ethnography is an inherentlyopen-ended practice. It is based on participation and observationin particular culturalarenasas well as acknowledgmentand employment of researcherreflexivity. That is, it relies heavily on "the acuity of the researcher-asinstrument"(Sherry 1991, p. 572) and is more visibly affected by researcherinterests and skills than most other types of research. Ethnography also uses metaphorical, of hermeneutic,and analytic interpretation data (see, e.g., Arnould and Wallendorf 1994; Spiggle 1994; Thompson 1997). Ethnographyis groundedin knowledge of the local, the particularistic, the specific. Although it is often used and to generalize, it is most often used to gain a type of particularizedunderstanding has come to be termed"grounded that and Strauss 1967). The rich qualitative knowledge"(Glaser content of ethnography'sfindings and the open-endedness that makes it adaptableto a variety of circumstanceshave led to its popularityas a method.This flexibility has allowed to ethnography be used for more than a centuryto represent and understandthe behaviors of people who belong to almost every race, nationality, religion, culture, and age group-and even behaviors of some nonhuman species groupings.Even with this impressivebody of ethnographic have work, however,it can be said thatno two ethnographies ever been conducted in exactly the same manner.This flexibility is one of ethnography's greatest strengths. Ethnographic methods have been continually refashionedto suit particular fields of scholarship,researchquestions, research sites, times, researcherpreferences,and culturalgroups.

consumerinteraction, new opportunitiesfor market-oriented it is also opening up opportunitiesfor marketingresearchers to study the tastes, desires, and other needs of consumers who interactin online communities. Marketingresearchersuse a variety of methods to study consumers. Qualitativemethods are particularlyuseful for revealing the rich symbolic world that underlies needs, desires, meanings, and choice (see, e.g., Levy 1959). Currently, the most popular qualitative methods are focus groups, personalinterviews,and "market-oriented ethnography" (Arnould and Wallendorf 1994). Although marketoriented ethnographyis an important techniquethat focuses on the behaviorof the people who constitutea marketfor a product or service, it is a time-consuming and elaborate method that requires considerable skill and substantial investmentsof researcherresources.Because it involves inperson researcher participantobservation, market-oriented ethnographyis also an intentionallyand unavoidablyintrusive method that precludesunobtrusive observationof naturally situatedconsumer behavior.Face-to-face focus groups (Calder 1977) and personalinterviews(Thompson 1997) are less time consuming, simpler,and more popularqualitative marketing research techniques than ethnographyis. However, their obtrusiveness, artificiality,and decontextualization of cultural marketing information are considerably greaterthan that of ethnography. This article extends the strengths of market-oriented ethnographyby demonstratinghow it can be efficaciously conducted online using existing online communities, often in an unobtrusivecontext. The novel, computer-mediated, textual, nonphysical, social cue-impoverished context of online communities may have hamperedthe rigorousinvestigation of these communities by researchers.Over the past several years, many anthropologists,sociologists, and qualitative marketingresearchershave writtenabout the need to specially adaptexisting ethnographicresearchtechniquesto the many cultures and communities that are emerging through online communications (see, e.g., Escobar 1994; Grossnickle and Raskin 2000; Hakken 1999; Jones 1999; Kozinets 1999; Miller and Slater2000). Although it does not break entirely new ground methodologically, this article addresses this importantneed by providingresearcherswith a rigorous methodology that is adaptedto the unique characteristics of online communities. or "Netnography," ethnographyon the Internet,is a new research methodology that adapts ethnographic qualitative research techniques to study the cultures and communities that are emerging throughcomputer-mediated communications. As a marketingresearchtechnique, netnographyuses the informationthat is publicly availablein online forumsto the identify and understand needs and decision influences of relevantonline consumergroups.Comparedwith traditional and market-orientedethnography,netnographyis far less time consuming and elaborate.Anothercontrastwith traditional and market-oriented ethnographyis that netnography is capable of being conducted in a mannerthat is entirely unobtrusive(though it need not be). Comparedwith focus groups and personal interviews, netnography is far less obtrusive, because it is conducted using observations of consumers in a context that is not fabricatedby the marketing researcher.It also can provide informationin a manner that is less costly and more timely than focus groups and personal interviews. Netnography provides marketing

This content downloaded from 137.43.71.99 on Tue, 30 Sep 2014 11:21:55 AM All use subject to JSTOR Terms and Conditions

Research for UsingNetnography Marketing
Although ethnographyis inherentlyan open-endedform of inquiry,ethnographerschoose from related field procedures and often confront similar methodological issues. Common ethnographic procedures that help shape researchers'participant observationinclude (1) makingcultural entree, (2) gatheringand analyzing data, (3) ensuring (4) trustworthyinterpretation, conducting ethical research, and (5) providing opportunitiesfor culture member feedback. Thorough accounts of these procedures exist for conductedin face-to-face situations(see, e.g., ethnographies Fetterman1989; HammersleyandAtkinson 1995;Jorgensen 1989; Lincoln and Guba 1985). However, networkedcomputing is a novel medium for social exchange between consumersthatchanges the particulars each of these research of level procedures,concomitantlyallowing an unprecedented of access to the heretoforeunobservablebehaviorsof interacting consumers. It is important,therefore, to provide a general description of the steps and proceduresinvolved in conducting netnographyas they are adaptedto these unique online contingencies. Although netnography, like ethnograis inherentlyflexible and adaptableto the interestsand phy, skill set of the individual marketingresearcher,these steps may act as a guide to researcherswho are interestedin rigorously applying the method to their own research. This combination of more rigorous online guidelines combined with an innate flexibility is novel, yet still faithfulto scholarly depictions of traditional ethnographic methodology (e.g., Fetterman1989; Glaser and Strauss 1967; Hammersley and Atkinson 1995; Jorgensen 1989; Lincoln and Guba 1985). After discussing these netnographicprocedures, I proceed to illustrate the richness of the technique with a short example of marketingresearchconductedin an online group devoted to the discussion of coffee. Entree.There are two initial steps thatmarketresearchers will find useful as preparation conductinga netnography. for First, researchers must have specific marketing research questions and then identify particularonline forums appropriate to the types of questions that are of interestto them. Second, they must learn as much as possible about the forums, the groups, and the individualparticipants they seek to understand.Unlike in traditionalethnographies,in the identification of relevant communities, online search engines will prove invaluable. Structurally,at least five different types of online community can be distinguishedthat may be useful to the conduct of market-orientednetnography(for more detail, see Kozinets 1999). First are boards, which function as electronic bulletinboards(also called newsgroups,usegroups,or usenet groups). These are often organizedaroundparticular products, services, or lifestyles, each of which may have importantuses and implications for marketingresearchers who are interested in particular consumer topics (e.g., McDonald's, Sony Playstation, beer, travel to Europe, skiing). Many consumer-orientednewsgroupshave more than 100,000 readers, and some have more than one million (Reid 1995). Currently,google.com has an excellent newsgroup search engine (acquiredfrom deja.com). Second are independentWeb pages as well as Web rings, which are composed of thematically linked World Wide Web pages. Web pages such as epinions (www.epinions. com) provide online communityresourcesfor consumer-toconsumer exchanges. Yahoo!'s consumer advocacy listings also provide useful listing of independentconsumer Web

63 pages. Yahoo! also has an excellent directoryof Web rings Third are lists (also called (www.dir.webring.yahoo.com). listservs, after the software program), which are e-mail mailing lists united by common themes (e.g., art, diet, music, professions, toys, educational services, hobbies). Some good search engines of lists are egroups.com and liszt.com. Finally, multiuser dungeons and chat rooms tend to be considerably less marketoriented in their focus, containing informationthatis often fantasyoriented,social, sexual, and relational in nature.Generalsearch engines (e.g., Yahoo! or excite) providegood directoriesof these communities.Dungeons and chat rooms may still be of interest to marketing researchers(see, e.g., White 1999) because of theirability to themes (e.g., certain industry, provide insight into particular demographic, or lifestyle segments). However, many marketing researcherswill find the generally more focused and more information-laden contentprovidedby the membersof boards, rings, and lists to be more useful to their investigation than the more social informationavailable in dungeons and chat rooms. In general, combining search engines (e.g., a World Wide Web search engine such as Yahoo! with a newsgroup search engine such as groups.google.com) will often provide the bests results for locating specific topics of interest. It is also importantto note that a broad and thorough computerizedsearch may be required,as the topic of interest may be categorizedat varyinglevels of abstraction, for example, at the brand,productcategory,or activity type level. When suitable online communities have been identified, the researchercan judge among them using criteriathat are specifically suitable to the investigation.In general, online communities should be preferred that have (1) a more focused and research question-relevantsegment, topic, or of group;(2) higher "traffic" postings;(3) largernumbersof discrete message posters;(4) more detailed or descriptively rich data;and (5) more between-memberinteractionsof the type required by the research question. These evaluations entail an importantadaptationof ethnographyto the online context, and their use distinguishesthe method of netnography from traditionalethnography.All the online forums (groups, rings, lists, dungeons,and rooms) may provideuseful access to people who are self-segmented by a certain type of lifestyle or market orientation, which researchers may, at their option, translate into private (one-on-one) online, real-time interviews (see, e.g., Hamman 1996). Before initiating contact or data collection, the marketing researchershould be familiarwith the characteristics (group behaviors, interests, and lanmembership,market-oriented guage) of the online communities. Data collection and analysis. After online communities are chosen, marketingresearchersare ready to begin collecting data for their netnography.There are at least two importantelements of this data collection: (1) the data the researchersdirectlycopy from the computer-mediated communicationsof online communitymembersand (2) the data the researchersinscribe regardingtheir observationsof the and meanings.As communityand its members,interactions, a distinct advantageover traditionalethnographers, netnographers benefit from the nearly automatic transcriptionof downloaded documents. With the addition of vastly lower search costs than face-to-face ethnography(particularlyin purely observationalforms of netnography),data are often

This content downloaded from 137.43.71.99 on Tue, 30 Sep 2014 11:21:55 AM All use subject to JSTOR Terms and Conditions

64

JOURNAL MARKETING FEBRUARY OF 2002 RESEARCH, rigorousnetnography using only observationand downloads and without writing a single field note. As dataanalysis commences (often concomitantwith data collection), the netnographermust contextualize the online data,which often proves to be morechallengingin the social Software cues-impoverished online context of netnography. solutions such as the QSR NVivo and Atlas.ti qualitative analysis packages can expedite coding, content analysis, data linking, data display, and theory-building functions (Paccagnella 1997; Richardsand Richards1994). However, classificationand coding of data are important concerns that involve trading off symbolic richness for coninevitably struct clarity (Van Maanen 1988). Perhapseven more than with ethnography, some of the most useful interpretations of netnographicdata take advantageof its contextual richness and come as a resultof penetrating metaphoricand symbolic interpretation (Levy 1959; Sherry 1991; Thompson 1997) ratherthan meticulousclassification. Providing trustworthy interpretation. For tracking the behaviorsof online communities,netnogmarketing-related the raphyis a stand-alonemethod. It is a way to understand discourse and interactionsof people engaging in computermediated communication about market-oriented topics. datacollection and analyDuringthe course of netnographic sis, the marketresearchermust follow conventionalprocedures so that the researchis reasonableor trustworthy (note that in most qualitativeconsumer research,the concept of is "trustworthiness" used ratherthan "validity," Lincoln see and Guba 1985; Wallendorfand Belk 1989). Netnographyis based primarilyon the observationof textual discourse,an important differencefrom the balancingof discourse and observed behavior that occurs during inperson ethnography (cf. Arnould and Wallendorf 1994). Informantstherefore may be presumed to be presenting a more carefully cultivated and controlled self-image. The uniquely mutable,dynamic, and multiple online landscape mediates social representationand rendersproblematicthe issue of informantidentity (Turkle 1995). However,netnographyseems perfectly suited to Mead's (1938) approach,in which the ultimateunit of analysis is not the person but the behavioror the act. I also draw insight from the work of the founder of "the linguistic turn" in philosophy, Ludwig Wittgenstein(1968), who might suggest that the posting of computer text is a social action (a communicative act or "languagegame"). If so, then every aspect of the "game" (the act, type, and content of the posting; the medium;and so on) is relevant observationaldata in itself, capable of being trustworthy. Using online data in this mannerrequires a radicalshift from traditionalethnography, which observes which observes and must recontexpeople, to netnography, tualize conversationalacts. This shift is necessary because the characteristics conversationin netnography differof are ent than they are in traditionalethnography:They occur through computer mediation, they are publicly available, they are generatedin writtentext form, and the identitiesof conversantsare much more difficult to discern. Generally speaking, links to fixed demographicmarkers can be useful for some marketingstrategy purposes (e.g., targeting),and netnographyis more limited than traditional ethnographyin this regard. The netnographermust determine the importance of these markers in relation to the researchquestionand to the authoritythatwill be grantedto is findings. It is worth noting that direct misrepresentation

plentiful and easy to obtain. In this environment,the netnographer'schoices of which data to save and which to pursue are importantand should be guided by the researchquestion and availableresources(e.g., the numberof online members willing to be interviewed,the ability of online members to express themselves, time, researcher skill). Dealing judiciously with instantaneousinformationoverload is a much more importantproblem for netnographersthan for traditional ethnographers. Because the online medium is famous (and infamous) for its casual social elements, messages may be classified first as primarily social or informationaland as primarily ontopic or off-topic (when the topic is the researchquestion of interest).Although researchersmight include all the data in a first pass or "grandtour"interpretation, they will generally want to save their most intense analyticalefforts for the primarily informationaland on-topic messages. The posters of online messages may also be categorized. Some novel categories for classifying them on the basis of their level of involvement with the online community and the consumption activity have been outlined by Kozinets (1999). Touristslack strong social ties and deep interest in the activity (they often post casual questions).Minglers have strong social ties but minimal interest in the consumption activity.Devotees have strongconsumptioninterestsbut few attachments to the online group. Finally, insiders have strongties to the online group and to the consumptionactivity and tend to be long-standingand frequently referenced members. For marketingresearchthat is useful for marketing strategy formulation,the devotees and the insiders representthe most important data sources. Preliminaryresearch reveals that devoted, enthusiastic, actively involved, and sophisticated user segments are representedin online communities by insiders and devotees (Kozinets 1999). It is also useful to note that online communities themselves tend to propagate the development of loyalty and (sometimes) heavy usage by socially reinforcing consumption. Therefore, marketing researchers interested in online word of mouth and influence may find it useful to trackhow tourists and minglers are socialized and "upgraded" insiders and to devotees in market-oriented online communities (Kozinets 1999). As with groundedtheory (Glaser and Strauss 1967), data collection should continueas long as new insights on important topical areas are still being generated.For purposes of precision, some netnographers may wish to keep close count of the exact number of messages and Web pages read (in practice, an extremely difficult measurement),as well as how many distinct participantswere involved. The strength of netnographyis its particularisticties to specific online consumer groups and the revelatorydepth of their online communications. Thus, interesting and useful conclusions might be drawnfrom a relativelysmall numberof messages, if these messages contain sufficient descriptiverichness and are interpretedwith considerableanalyticdepth and insight. A time-tested and recommendedway to help develop this insight is to write reflective field notes. In these field notes, recordtheir own observationsregardingsubnetnographers texts, pretexts,contingencies, conditions,and personalemotions that occur during the research.These written reflections often prove invaluableto contextualizingthe data and are a recommended procedure.However, in a sharp break from traditionalethnography,a researchercould conduct a

This content downloaded from 137.43.71.99 on Tue, 30 Sep 2014 11:21:55 AM All use subject to JSTOR Terms and Conditions

for Research UsingNetnography Marketing discouraged in most online forums. Codes of etiquette (see Gunn 2000) and other social pressures are often in effect. Misrepresentingoneself as a member of a restrictedgroup (e.g., women only, under 18 years of age) is an offense punished by flaming, ostracism, and banishment.However, triangulation of netnographicdata with data collected using other methods, such as in interviews,focus groups, surveys, or traditionalin-person ethnographies,may be useful if the researcherseeks to generalize to groups other than the populations studied. Generalizing the study beyond particular online groups may not be necessary.Yet careful triangulation and long-term immersionin the communitycan be useful to help marketingresearchersdistinguishhard-core,marginal extremists from a more typical group of consumers.It should be noted that, just as during in-person exchanges, extremistsare derided. In the largercommunities(with hundreds of active members) moderateviews seem to prevail. Online communities present fairly explosive environments and, freed of many of the usual social restraintsemployed during in-person gatherings, hard-coreextremists are often soundly condemned. In summary,throughoutnetnographicdata collection and analysis, marketingresearchersmust be conscious that they are analyzing the content of an online community'scommunicative acts ratherthan the complete set of observedacts of consumers in a particularcommunity.This is a crucial difference between netnographyand traditionalethnography. Stories of online misrepresentation legion and important. are Generalizations to markets or communities other than the one studied, online or offline, must have corroborating evidence. To be trustworthy,the conclusions of a netnography must reflect the limitations of the online medium and the technique. Research ethics. One of the most importantdifferences between traditionalethnographyand netnography may be in issues of researchethics. Marketingresearchersdesiring to use netnography a methodare obliged to considerand folas low ethical guidelines. These guidelines for ethical social science researchin cyberspace have been the topic of recent debate. Ethical concerns about netnography turn on two nontrivial, contestable, and interrelated issues: (1) Are online forumsto be considereda privateor a public site? and (2) What constitutes "informedconsent" in cyberspace?A clear consensus on these issues, and thereforeon ethically has appropriate proceduresfor netnography, not emerged. In a majordeparture from traditional face-to-facemethods such as ethnography,focus groups, or personal interviews, netnographyuses informationthat is not given specifically and in confidenceto the marketing researcher. consumers The who originally created the data do not necessarilyintend or welcome the data'suse in researchrepresentations. NetnograThe uniquely unobtrusive phers are professional "lurkers": natureof the methodis the sourceof muchof its attractiveness and its contentiousness.If marketing researchers undertaking act and netnography in a mannerfoundto be irresponsible disrespectfulby consumers,they may well damagethe medium (by either suppressingoutrightor drivinginto secrecy previously open social interactions)and thereby "poisoning the researchwell" (Reid 1996). This is a realrisk.White(1999, p. B I) reportshow music promotersavoided identifyingthemselves when they acted as both online marketers marketand ing researchers"tryingto get a quick gauge on something, where you don't want anyone'sguardto be up."

65
There is genuine debate about the public versus private about the electronic eavesdropissue. Speaking particularly of observationalethnography,Rafaeli (quoted in Sudping weeks andRafaeli 1995) summarizesthe consensus of a certain group of scholarswho debatedthe privateversus public issue by statingthat informedconsent was implicit in the act of posting a message to a public area.Given thatcertainprecautions were taken to provide anonymity to informants, this group of scholars approvedan ethical policy in which the informed consent of Internetposters was not required. King (1996), however,bases his analysis on the notion that online forumsdissolve traditionaldistinctionsbetween public and private places, making conventional guidelines of anonymity, confidentiality, and informed consent unclear. King (1996) therefore concludes that because consumers might be deluded about the quasi-public nature of their ostensibly private communications, gaining additional informed consent from them was the responsibility of researchers.Sharf (1999) echoes this heightened sensitivity to the ethics of even observationalnetnography. The potential for netnographyto do harm is a real risk. For example, if a marketingresearcherwere to publish sensitive informationthat was overheardin a chat room, this or might lead to embarrassment ostracism if an associated person's identity was discerned (see Hamman 1996). Several informantshave requestedthat I not publish statements they have posted on public bulletin boards, even though I always guarantee their anonymity. I have always honored these requests. This evidence supports the contention that "thereis a potentialfor psychological harmto the members of these [online community]groups, depending on the way results are reported" (King 1996, p. 119). Researchers who have published cultural secrets; poror trayedpeople and practicesinaccurately; treatedcustoms, individuals,and beliefs disdainfullyhave taintedthe history of ethnography.The same potential for harm exists for In netnography. a time of increasingpublic scrutinyof coractions and computerprivacy issues, as well as instiporate tutional review board scrutiny in academia, netnographers would be wise to consider the chief ethical concerns apparent in netnography: of privacy,confidentiality,appropriation others' personalstories, and informedconsent (Sharf 1999). Therefore,I recommendfour ethical researchprocedures for marketingresearchersusing netnography. Although they parallel practices in conventional ethnography,these first three proceduresare not at all obvious to people who are used to conducting Web searches and Internet research. should fully disclose They are as follows: (1) The researcher his or her presence, affiliations, and intentions to online community members during any research; (2) the researchersshould ensure confidentialityand anonymity to informants;and (3) the researchershould seek and incorporate feedbackfrom membersof the online communitybeing researched.The fourth procedureis specific to the online medium:(4) The researchershould take a cautious position on the private-versus-public medium issue. This procedure requiresthe researcherto contact community membersand obtain their permission (informed consent) to use any specific postings that are to be directly quoted in the research. Permission must be obtained for using idiosyncraticstories as well (see Sharf 1999, pp. 253-55). Before using any online artifacts,such as newsletters,poetry,stories, or photographs, permission from the copyright holder must be

This content downloaded from 137.43.71.99 on Tue, 30 Sep 2014 11:21:55 AM All use subject to JSTOR Terms and Conditions

66

OF 2002 JOURNAL MARKETING RESEARCH,FEBRUARY tectonic shifts in the coffee marketin the pastdecade. Major consumer packaged goods companies such as General Foods and Procter & Gamble were apparently caught unawareby the Seattle coffeehouse trend that came to be personified by the "Starbucksinvasion"that overtook boutique coffee shops and subsequentlyencroached on supermarket aisles (see Pendergrast 1999; Schultz and Yang 1999). Starbuckssimultaneouslyraisedthe consciousness of coffee connoisseurship, the demand for coffee shops, the sales of coffee-flavored ice cream and cold drinks, and the marketprice of a cup of coffee. An understandingof coffee meanings can be gleaned from a netnography a dedicatedcoffee group.As with the of communities, membershipof many online market-oriented the members of this coffee group can be characterizedas devoted, enthusiastic, knowledgeable, and innovative. In theirenthusiasm,knowledge, and experimentation with new forms of coffee consumption,they can provide information similar to that from "lead users," the inventive consumers who are at the leading edge of significant new marketing trends (von Hippel 1986, 1988). Although some may be marginalor hard-coreusers, theircreativeideas and insights should not be discounted as without value. By carefully evaluatingtheir innovativeideas and by cross-validatingthe quality of informationthey provide aboutcurrentconsumption trends with other information sources, this study reaches conclusions that can inform decisions by members of the coffee market, such as consumer packaged goods companies, coffeehouse retailers,coffee mail-ordercompanies (both online and offline), and advertisersworking on coffee-related accounts. By carefully corroborating,interpreting, and critically evaluating this information, this researchcould generateinsights to informnew productconcepts, positioning strategies, advertisingcampaigns, distribution tactics, and other marketingstrategiesand practices. Understandingthis online community's messages and its mediumcan also provideinsight into the use of newsgroups and other online media for coffee-relatedmarketing. Entree in online coffee culture. This netnography into online coffee culture began with an overview of the newsgroups that contained the term "coffee" and were available from my local server.These revealed three potential newsgroups, , , and ,as well as severalothers. I chose because it had by far the highest amountof traffic(approximately 75 messages per day) and therefore contained the most data.According to 1995 Arbitrondata, is ranked 1042 out of all newsgroups,is carriedby 40% of all service providers,and is read by 55,939 people worldwide (Reid 1995).2 It contains a core of insiders who are frequently quoted and referenced by other community members, deferred to by existing and new members, and mentioned by members as importantarbiters of coffee taste. Therefore,from an informaltype of networkanalysis, these insidersseem to be usefully conceptualizedas opinion leaders in the local context of this particular online community. The community also contains many minglers, who stay on for periods of six months to a year, and a large numberof
2Given the growth of the Internetbetween 1995 and 2000 and the doubling of message postings on duringthatperiod,it is likely that as of 2000, the newsgrouphad more than 100,000 readersworldwide.

granted. Following these specially adapted research techniques will help ensure that ethical netnography is conducted that avoids poisoning the well for futureresearchers. Member checks. A member check (Arnould and Wallendorf 1994, p. 485; Hirschman 1986, p. 244; Lincoln and Guba 1985) is a procedurewhereby some or all of a final research report's findings are presented to the people who have been studied in orderto solicit their comments. Member checks prove particularly valuable for three reasons related to the dissimilarity of netnographyfrom traditional ethnography. First, because they enable researchers to obtain and elicit additional,more specific insights into consumer meanings, they are particularlyvaluablefor conduct(i.e., member ing an unobtrusive,observationalnetnography checks providethe opportunities addeddevelopmentand for error checking). Second, they help ameliorate some of the contentious ethical concerns described in the previous section, while still preservingthe value of unobtrusiveobservation (because member checks are usually conducted after data collection and analysis has concluded). Third,and perhaps most important,memberchecks can help establish an ongoing information exchange between marketing researchersand consumer groups that is unprecedentedin traditionalqualitativeresearch.Indeed, using the conduct of netnographyas a forum for ongoing, widespread, bidirectional communicationbetween organizationsand theircommunities of customerscould help realize some of the hidden potential in the paradigmof relationshipmarketing. As distinct from face-to-face ethnography (in which member checks are burdensomeand onerous and therefore are sometimes omitted) and focus groups and interviews(in which member checks are not usually employed), netnographic memberchecks are a generally simple and convenient matter.The low costs of computer-mediated communication enable the marketingresearcherto easily provideany interested reader with some or all of the research text, by either posting it on a Web page or sending it as an e-mail attachment.The elicitation and collection of informant comments is also greatly simplified and expedited through email. Because memberchecks, as well as the otherelements of netnography, can generally be completed in a more timely mannerthan face-to-face market-oriented ethnography, they provide marketersthe opportunityto detect and respond more quickly to the changing consumer tastes, meanings, and desires that underlie important marketing trends. Given these methodological considerations, I now proceed to a brief illustrativeexample of market research using netnography. ILLUSTRATION: OF ANALYSIS THEMEANINGS OF CONTEMPORARY COFFEE CONSUMPTION AN IN ONLINECOFFEE COMMUNITY ApplyingNetnographicMethodology In the shortsection thatfollows, netnographyis illustrated as a marketing research method. Netnography is used to explore and analyze some of the meanings and symbol systems that surround contemporarycoffee consumption (in those surrounding particular, espresso and Starbucks)for the posters to an online community that is dedicated to coffeerelated discussion. Understandingand trackingthese meanings and symbol systems are of considerable practical importance.As many marketersare aware, there have been

This content downloaded from 137.43.71.99 on Tue, 30 Sep 2014 11:21:55 AM All use subject to JSTOR Terms and Conditions

for Research UsingNetnography Marketing tourists,who come and go with specific queries.Priornewsgroup surveys indicatethatpostersare mostly male and well educated, with an average age of 48 years. As part of ongoing research,I followed and related newsgroups and downloaded noteworthy messages starting in February1998. I read several hundredmessages over the 33 months of netnographicresearch. In addition, the research was informedby searches of coffee-relatedWeb pages, Web rings, and mailing lists; books about coffee; coffee consumption experimentation; and in-person product-related discussions with coffee consumersand connoisseurs.Limiting the investigationto 179 postings that I downloadedand printed kept the amount of data limited to a manageable level. The majorityof the messages thatI downloadedwere posted between July and November 2000. Data collection. The 179 postings were preclassified (before downloading) into topics that were eitherrelevantor not relevantto the researchtopic of interest(contemporary coffee meanings). For example, threads(a threadis a set of interrelatedbulletin boardpostings) such as "Coffee Poem" and "How to make a great cappuccino at home" were pursued. Threads such as "NY Chocolate Show" were not, because they were judged not to be relevant. Several message threads related to Postum, such as "Anyone tried or heardof this?"were exploredand downloaded,which elucidated what constituted both good coffee and its antithesis. As the investigation narrowedonto discussion of Starbucks, the "Weird Starbucks Experience,""Peets So Good," and "Americans-your thoughts on Starbuckswanted"threads were downloaded. The importanceof espresso to the community was also evident as the investigationnarrowed.This topic was explored in "Woohoo,just got my Silvia/Rocky." These threads were chosen for their rich content, descriptiveness, relevanttopic matter,and conversationalparticipation by a range of differentcommunitymembers.The range of conversational participationwas importantto avoid the researchbeing misled or undulyinfluenced by a minorityof and unrepresentative vocal extremists. is Using carefully chosen message threadsin netnography akin to "purposivesampling" in market-oriented ethnography (Lincoln and Guba 1985; Wallendorfand Belk 1989). Because findings are to be interpreted terms of a particuin lar sample, it is not necessary for the sample to be representative of other populations. However, there is the potential for anonymous self-promotion by manufacturers retailand ers. Therefore, messages that were suspect in this manner (i.e., overly engaging in promotionor containingan e-mail address related to the company on which they were commenting) were excluded from the data set. In addition, and when it was possible to do so, apparentlyoff-topic useless talk was coded and excluded from analysis because it did not pertainto the centraltopic of coffee consumption. Analysis and interpretation.The coding of the postings involved both data analysis and data interpretation (Spiggle 1994, p. 492). Netnographicdata in each categorized interaction were compared with the data from other events that were coded as belonging to the same category,and theirsimilarities and differences were examined (Glaser and Strauss 1967; Spiggle 1994). Each category later formed a theme, abstractor groundedtheory,or "metaobservation" (Arnould and Wallendorf 1994; Lincoln and Guba 1985; McCracken 1988). For this research,the volume of text was 198 double-

67 spaced 12-point font pages, representing117 postings containing 65 distinct e-mail addresses and user names (likely related to the numberof people posting messages). Disconfirmingevidence was sought, both within the data set and in later searchesof Web pages and the newsgroup, and resulted in several early themes being rejected. Concomitantly with analysis, the data were subjected to interpretation,which, as Spiggle (1994, pp. 497, 500) describes it, is "playful, creative, intuitive, subjective, particularistic, transformative, imaginative,and representative." Ethics and member checks. To ensure research ethics, I identified myself in postings to the community, told members about the observation,and provided my credentials. I sought permission to use direct quotations, and it was (Lincoln and granted.To ensurea trustworthy interpretation Guba 1985), I conducted member checks with nine online informants. Member check informants said they were "impressed"by the netnographyand thought it was "perceptive"and even "fantastic." They also had several suggestions. Member checks resulted in revisions to the depiction of basic coffee (including press pot and vacuum pot preferences), commodification, and religious devotion and the provision of some additionalgroup characteristics. A Brief Netnographyof Online Coffee Cultureon the Usenet Newsgroup As Sherry(1995, p. 356) has noted, "Coffee is among the preeminent vessels of meaning in consumer culture" (see also Pendergrast1999). This richness of meaning is evident in the vital and virtuosic exchanges that transpirethrough . Like the members of any thoroughgoingculture, the denizens of the newsgroupspeak their own language.Theirposted conversationsare pepperedwith terms that are unfamiliar to the uninitiated:baristas and JavaJocks,cremas and roastmasters, tampersand superautomatics, livias and tiger flecks. It is the specialized language of the coffee lover, conveying many of the subtleties of coffee taste and preparation. Understandingthe language of consumer segments and its specific underlyingsocial motivationsis a key aspect to achieving the marketorientation(Kohli and Jaworski 1990) that can successfully conceptualize new products, employ existing and new channels, and write potent advertisingthat meaningfully communicates to markets. Although a full translationof this newsgroup's language is impossible in this article, I examine some importantculturalthemes contained within it. In this short netnography, examine themes I of distinction, consumption webs, commodification concerns, and religious devotion. I specify marketingresearch implicationsthroughoutand extend them in the conclusion. Distinction: decoding the language of motivation. On , the specifics of coffee connoisseurship are repeatedlytaught.One of the first things taughtis that basic coffee, the type that most people enjoy in their offices and homes, is usually beneathcontempt because it is "normally very badly prepared and stale." Proper coffee, flavorful coffee must be preparedcorrectly. This means avoiding paper filters and drip coffee (and percolators) and instead using gold filters, cafetiere, press pots, or vacuum pots (in orderof preference).Althoughespresso may not be the most frequentlyconsumed form, it is the most discussed form of coffee on the newsgroup. Real coffee, precious coffee,

This content downloaded from 137.43.71.99 on Tue, 30 Sep 2014 11:21:55 AM All use subject to JSTOR Terms and Conditions

68

2002 OF JOURNAL MARKETING RESEARCH,FEBRUARY
The marketingresearchimplicationsof these postings lay in the way some coffeephiles describe their motivation to develop taste, which led them to spend large amounts of money on coffee equipment.When they are acculturatedto the propertaste of espresso and its rarity,these consumers rejectconventionalcoffee offerings (often giving them terrible, excretorynames) and popularcaf6s (often emphasizing their robotic qualities) and are drawn into multiple investments to which thereseem no end. The previouscomments, in which a coffeephile ascribes his increasinginvestmentto the influence of a fellow newsgroup member, suggest the power of the newsgroup to acculturateconsumption practices. This acculturating force, which drives increasing investmentsin a new cultural interest,has been termed the "Diderot effect" (McCracken 1990). In the newsgroup,there is evidence for an acculturatedtransition from regularhome-brewedcoffee to basic press pots (such as the Bodum) to betterpress pots to vacuumpots. Another is from home brews to caf6-bought coffee to caf6-bought "fancier" drinks such as latti and cappuccino to storebought espresso to homemade espresso, which requires a starter machine, then a better machine, a coffee bean grinder,then a coffee bean roaster,then a kitchen vent for the roaster,then better beans, and so on. This subtle inculcation of coffee tastes (on a trajectoryculminatingin a taste for espresso) is often mapped out in coffeephile communications, tracinga gustatoryroute through,for example, cappuccino, macchiato,and con pannasto espresso. In total, this set of united productscan be interpreted a as "productconstellation"(Solomon and Assael 1987) that is linked to the real or desired social class of these coffee drinkers.For marketingresearchers,this productmap might be thoughtof as a particular consumptionweb that increasdrawsa groupof consumersinto deeperand more proingly found levels of (sub)culturalinvolvement and enthusiasm, the consumption,and investment.Understanding configuration of these particularconsumption webs would provide coffee-related manufacturersand retailers with ideas for new product and service offerings and bundling (e.g., bundling together brandsof products that are perceived as associated with other brands, bundling together kitchen venting systems with roasters,bundlingfeatureson espresso machines to produceconsumer-related forms of coffee). brands: brand image and communityconCommodified cerns. Another importantculturalcode links good coffee to passion, artistry,and authenticityas a fully realized human being. The discussions that reveal this code centeredon the nature of the barista, or coffee server. The online coffeephiles proclaim that "the product(be it food or coffee)" is because it always "anexpressionof the maker'spersonality" is "anart afterall" (Vincent,3posted on August 6, 2000), that"barista" implies "anartisan... like a seasoned sommelier or vintner" (Angelo, posted on August 9, 2000). Several posters claim that they would not visit a caf6 whose baristaswere not coffee lovers (and several othersdisagreed).An existential dimension is added by one of the originalposters,in which he rejectsthe term"artisan" but says that being an authenticbarista"hasto do with the way you live your life." Coffee becomes, to this culture member,a metaphorfor life, in which either life is mere rule are 3Pseudonymns used throughoutto protectinformantconfidentiality.

essential coffee (both literally and figuratively)is espresso, consumed without "cow juice" or sugar. Making good espresso is a complicated affair. It involves careful attendance to the water, the grind, the timing of the shot, knowledge of the machine, a clean portafilter(portablefilter) and the screen, the tamper,the blend, the ambient temperature, age of the coffee, the degree of the roast, the air humidity, incoming water temperature, internal boiler temperature, and even such mystical elements as the mood of the barista [coffee server] and "good old-fashionedluck." These are not merely functionalconsiderationsbut online incantationsof status, upwardsocial movement, and hedonism thatare intendedto manifestand demonstratethe "distinction"or "culturalcapital"of upper-classtastes and abilities (Bourdieu 1984; see also Holt 1998). There is an elitist or classist "snobappeal"to coffee knowledgethat motivates discerning tasting, as well as the reading of coffee-related books such as UncommonGrounds(Pendergrast1999) and authoritativeguides such as the site of David Schomer (http://www.lucidcafe.com/cafeforum/schomer.html). As Levy (1981) convincingly demonstrates, there are strong links among discernment,social class, and the acculturated sense of taste. This acculturation the complexities of taste of and flavor appearsto transpireonline. For example, the flavor of good espresso is much discussed and describedonline (it is not too watery and not too burnttasting but has a slight agreeable bitternessand a slight astringency). Also, the group's discursiveactions enact a deep desire to go behind the scenes; to understandwhat it is that makes a particular type of coffee superior;and then to capture,reproduce, and by reproducingbecome a part of the productive consumption of the experience. This productiveconsumption is also a status marker.Home espresso brewing is a fairly expensive hobby (but not prohibitivelyso for the U.S. middle class), which is partiallywhy it can serve as a distinctive marker.This need not only to consume but also to actively produceis a hallmarkof deep devotion to a particular consumption orientation,such as is found in a range of subcultural,sports, music, and media fan experiences (see, e.g., Fiske 1989). Consumptionwebs: mappingthe paths of desire. The key to these descriptions is not merely their specifics (though these are equally importantto consumersand the marketers that seek to serve them) but the amazing raritythat is conveyed within them, the scarcity evident in all the stressing over when to pull, when to tamp, how to time, which machine, which coffee bean. One member cautioned that only one of every five pulls is worthdrinking,which makes educating the palate about good espresso a difficult task. As with wine production and tasting, productionand discernment of espresso takes time and practice.Some coffeephiles claim that their taste bud trainingtook months. One active poster stated thatthe trainingperiod lasted nine months.This coffeephile noted that the downside to educating his palatewas that he became a slave to coffee and eventually spent huge amounts of money to keep himself from being subjectedto more ordinarycoffee (which had become unbearableto him). He also noted that there was no end to his involvement. Once acculturated,he kept finding new pieces of coffee equipment that he could not live without, a state of affairs he jokingly yet pointedly blamed on his fellow coffeephiles.

This content downloaded from 137.43.71.99 on Tue, 30 Sep 2014 11:21:55 AM All use subject to JSTOR Terms and Conditions

for Research UsingNetnography Marketing following "or you really experience what life is all about" (Vincent, posted on August 9, 2000). The mark of authenticityis baristaswho "drink/livecoffee" just as do the denizens of . It is passion that matters: "Coffee is the passion of a baristaand a lifelong profession" (Peter, posted on August 13, 2000). This emphasis is also present in an online debate among Starbucks employees. The more passionate coffee drinker(former employee) accused the other (currentemployee): Coffee is just another for product you too. Youcould twaddlers flame or just as well be sellingthoseturnip retardant condoms,butas long as you are havingfun and payingyourbills, thatis all thatmatters you, to right?I am afraidthatit is not quite thatsimple for and manyof us. Wetakeourcoffee veryseriously, to have it demeanedin such a manner a slap in the is face. Coffeeis muchmorethana tool. It is passion,it is intrigue,mystery,seduction,fear, betrayal,love, emotionthatyou can hate,andany othercore human think of, all wrappedinto one little bean. (Peter, postedon August14, 2000) Because Peter's rich and revealing comments were applaudedand referencedby manydifferentmembersof the online community, they seem to cut to the core of some important (and shared) impressions of Starbucks among members. Coffee is emotional, human,deeply and personallyrelevant-and not to be commodified(Kopytoff 1986) or treatedas "justanotherproduct." This concern is reflected in two negative newsgroupnicknamesfor Starbucks: As an expensive and faceless corporateentity, it is "*$"; as a killer of mom-and-poplocal stores, it is "corporate coffee." Presentingimportant culturalclues to the positioning of any new coffee marketerthat seeks to compete with the Starbucksbrand,the discussion of Starbucksturned into a more general discussion of the perils of commercialization and cultural commodification.The resentmentover the commodification of coffee connoisseurship leads to dialectics of authenticityand genuineness: WhatI am comingto in my own life and consumer is behavior thatI wantto support savor truespeand the eat cialtyitemswhileI can.I'drather Barry's fudge... thanGodiva "fauxspecialty" chocolates. I'drather And drinkthe local caf6's coffee ratherthan Starbucks's because, well, those tiny, passionate companiesare moreprecious thanStarbucks.... corporation with Any food chemistscan makeStarbucks's IMO[in product, driven romantic would my opinion]. Onlya passionate, coffeeday in andday keepmaking top-notch specialty out. Lose Starbucks another and clone clicks into that economiceco-niche.Lose a loveror a heroand you might wait a long time until anothercomes along. November 2000) (Fred, 19, postedon Fred'sdialectictranscends functional characteristics such as coffee flavor.Its overriding themeis thatvendorsor manufacturersshould demonstratea genuine passion for the product equal to, or close to, that of its connoisseurconsumers.This sentimentresists,in some sense, the commodification labor of in which people can be mechanisticallytrainedto produce items withoutenjoyingthemas consumers.It is a postmodern longing to return to productive consumption (Firat and Dhalokia 1998).Fred'sdialecticof commodification reflectsa search for authenticity, to the local, caring by producers, ties

69 and In craftsmanship, artistry. the same posting,Fredexplains thatto support Starbucks not to support is local merchants such as Tom,a coffee "maven" who is obsessedwith"theZen of the cup" (a spiritual-religiousmetaphorconnotingdevotion and To authenticity). supportlocal caf6s is a statementnot only aboutcoffee but abouthumanvalues and the world.As Fred states,it helps maintain"a worldof beautyand passion." Religious devotion: uncovering meaningful metaphors. This utopian"worldof beauty and passion"is evident in the wonderfully detailed accounts of coffee preparationand consumption provided in the newsgroups, which serve as sources of espresso education, expressionism, and exhibitionism. Members draw one another in with dramaticflair and literarydevices that playfully hint at the joyful mindset of the coffee connoisseur and, tongue-in-cheek, employ sacred metaphors.Describing himself in the third person, Jerry lovingly details (in several pages of text) his exact experiences with his new coffeemaker: Hehitthebrewswitch[onhisnewLivia90 cappuccino/ At Then..., beautiful espressomaker].... first,nothing. reddish-brown crema... the"tiger flecks" hadheard he so muchaboutbutrarely seenflowedforthandfell had just shortof two ouncesin 25 seconds.He stoodjust the a admiring cremawhen suddenly voice called to him, "The milk! The Milk!" (Jerry, posted on November 2000) 2, As with Fred's "passion,"his David and Goliath-story "hero,"and his "world of beauty,"the language Jerry uses here is romantic,idealistic, and biblical. The crema(oil from the coffee beans) is "beautiful," it "flowed forth"much and like a river of milk and honey might do for Old Testament Israelites.Jerrydid not simply rememberto steam the milk but portrayedit as "a voice" that"called[un]to him,"as if he were a biblical prophet.The dramaand religion may be parodic, but they are repeatedly present and meaningful as a local cultural code, indicating that this is not merely the meanderingof extremists.For example, otherpostings replicate the dramaticand religious metaphor,calling the lack of passion by a "Starbucks jock" "Sacrilege!"and the addition of sugarto espresso the markof a personwho "has no soul." The interpretive coup de grace may be in the termthatthis community of coffeephiles uses for the elusive, religious experience, the exhaustiveapotheosis of espresso moments, the holy grail of the coffee dream quest. It is called a "god shot."It representsthe sublime momentof coffee productive consumption,an absolutely perfect, indefinablemoment of glory, one that cannot be captured, reproduced, or summoned at will. A god-shot is a supernatural event. It is a momentwhen humanbeing and natureare reunitedin a perfect convergence of elements (water, fire, air, and earth/ grounds), resulting in a perfectly pleasurable occurrence. This interpretation does not suggest thatcoffee consumption is actually a religion for these coffeephiles. But for them it has religious aspects of search, passion, and transcendence (see Belk, Wallendorf,and Sherry 1989) and deeply meaningful ties to identity (Fiske 1989). As comments to Jerry's postings indicate, these metaphorsare highly motivational and persuasive and thus are of interest to marketing researchers.

This content downloaded from 137.43.71.99 on Tue, 30 Sep 2014 11:21:55 AM All use subject to JSTOR Terms and Conditions

70
AND IMPLICATIONS CONCLUSION

FEBRUARY 2002 OF JOURNAL MARKETING RESEARCH, by Starbucks.New brandsand blends of beans, new means of deliveringthe freshestof fresh beans (online and offline), new means of roasting, new bean roasting services, new espresso and cappuccinomachines, new forms of education and instruction,new coffee tasting clubs, and new types of cafes are premium opportunitiesthat await furtherevaluation and explorationby opportunisticnew productdevelopers and marketeducators. From the practical standpointof professional marketing online communities for researchers,identifyingappropriate particularmarketingresearch clients is more art than science. As this netnographydemonstrates, the informationpresentin a particular newsgroupis likely to be of more value to certain types of industry players. In , the information is particularly valuable to online and offline marketersof high-end espresso makers, roasters, grinders, caf6s, and roasted and unroastedcoffee beans and others that sell coffee connoisseurship-related goods. However, the information provided in the netnography about coffee's cultural cachet (relating it to social distinction, artisanship,craftsmanship, personalinvolvement,passion, authenticity,humanity,and religious devotion) might be useful in articulating range of positioningand branding a strategies with wider appeal. For example, newsgroupparas ticipants'critiqueof Starbucks'sbrandmeaning(regarded mechanistic, dispassionate, oppressive, overly large, and lacking humanityor a human touch) might be perceived as feedback to Starbucksand an opportunityfor Starbucks's competitors. If the Starbucksbrand is becoming passe, a mere symbol ("*$") of overroasting,a good place to read and hang out but not to drink coffee, then the next generation of coffee brands to tap into the discriminatingcoffee ethos will likely thriveby positioningon the oppositeend of these dimensions: human, passionate, roasted right, free, alive, locally involved, existentially complete. These culturalmeanings will draw on rich associationsto artand artisanship, craftsmanshipand connoisseurship-perhaps even religion and spirituality-and do it in a manner that is authenticand genuine. Coffee companies with a true market orientationwill find opportunitiesin this netnographic data and their own coffee consumercommunionnot simply for a new appearanceor faqade but for a depth of marketplace involvementand the understanding a genuine, passionate of coffee lover. Conclusion Online communitiesdevoted to consumption-related topics are an increasinglyimportantsource of data for marketing research.These groups may be construedas individual marketsegments that are of interest in their own right and may be of noteworthysize. As purchaseand consumption decisions are discussed and debated in online communities, it is importantthat marketingresearchershave rigorousand ethical methodological procedures to collect and interpret this data in this novel and challenging context. As the illustrationdemonstrates,netnographycan be a useful, flexible, ethically sensitive, and unobtrusivemethod adaptedto the purposeof studyingthe language,motivations,consumption linkages, and symbols of consumption-oriented online communities.

Deriving from naturally occurring, communal, crossconsumer interaction that is not found in focus groups or personal interviews, netnography reveals interesting consumer insights, impressions, linguistic conventions,motivations, consumption web linkages, and symbols. It provides feedback on brands and productsthat has not been elicited in any way by marketers,eliminatingthe researcher-induced demand effects of these methods and of traditionalethnographic inquiryand interview.The method achieves all this in a mannerthat is far more unobtrusive,convenient, and accessible than traditionalethnography.It is also far more economical. As the consumer verbatims and descriptions provided previously may attest, online consumers tend to be knowledgeable and educated and provide interestingconsumption insights. Because message posters are in some respect selfselected for their eloquence, the data they provide can be extraordinarilyrich. Online posters appear to spend large amounts of time and money on their focal consumption activity. By carefully evaluatingtheir innovativeideas, their knowledge base, and their consumer insights, marketing researchers can obtain useful information similar to that obtainedfrom lead users (von Hippel 1986, 1988). Ideas for innovative trends in particularrealms of consumptionsuch as novel productconcepts may thus be initiatedby investigations that begin with netnography.However,careful consideration and cross-validation of the online data will be critical to the researcher'savoiding being misled by overly zealous or vocal community members. Similarly, crossvalidationand a careful categoricalanalysis will be required for an understanding the relationshipof differenttypes of of online community members to typical online and offline consumers. Implicationsof Netnography Given the familiar diffusion of innovationsmodel, it can easily be arguedthat today's devoted or extremeconsumer's perspectivecan yield importantinsights into the more mainstreamconsumer behaviorof tomorrow(von Hippel 1988). The implications of this marketingresearchfor wise coffee marketersare thus considerable. It may have appeared, in the wake of Starbucks,that marketershad been one-upped by the Seattle coffeehouse craze and had missed the opportunity to raise the marketto its new upscale level. Yet if the market intelligence of the group is correct, coffee marketers have barely even begun to plumb the depths of taste, status, and snob appeal thatare waiting to be explored by discriminating coffee consumers who are in need of marketeducation. Experimental and innovative online coffee consumers offer a range of discoveries that, like a lead user analysis, inform the understandingof coffee marketingtrends. Not only does offer the enticing consumptionwebs and socialization pressures that can turn decaffeinated drinkersinto home-roasting,home-brewing,espresso savorers, willing to throw out four shots of expensive brew in search of the all-elusive but sublimely satisfying "god shot," but it also suggests that there is far more to coffee consumption than the in-personsocial, communal, and socially responsibleaspects that have been so successfully exploited

This content downloaded from 137.43.71.99 on Tue, 30 Sep 2014 11:21:55 AM All use subject to JSTOR Terms and Conditions

Research for UsingNetnography Marketing
REFERENCES Almquist, Eric and Kenneth J. Roberts (2000), "A 'Mindshare' Manifesto,"MercerManagementJournal, 12, 9-20. Armstrong,Arthurand John Hagel III (1996), "The Real Value of On-line Communities,"Harvard Business Review, 74 (May/ June), 134-41. Arnould,Eric J. and Melanie Wallendorf(1994), "Market-Oriented Ethnography:InterpretationBuilding and Marketing Strategy Journal of MarketingResearch, 31 (November), Formulation," 484-504. Belk, Russell W., Melanie Wallendorf, and John F. Sherry Jr. (1989), "The Sacredand the Profane:Theodicy on the Odyssey," Journal of ConsumerResearch, 16 (June), 1-38. Bourdieu,Pierre(1984), Distinction:A Social Critiqueof the Judgment of Taste,RichardNice, trans.London:Routledge& Kegan Paul. Bulik, Beth Snyder (2000), "The Brand Police," Business 2.0, (November 28), 144-55. Calder, Bobby J. (1977), "Focus Groups and the Natureof Qualitative Research,"Journal of MarketingResearch, 14 (August), 353-64. Escobar, Arturo (1994), "Welcome to Cyberia: Notes on the Anthropology of Cyberculture," Current Anthropology, 35 (June), 211-31. Fetterman,David M. (1989), Ethnography:Step by Step, Vol. 17, Applied Research Methods Series. Newbury Park, CA: Sage Publications. Firat,A. Fuat and Nikolesh Dholakia (1998), ConsumingPeople: From Political Economy to Theaters of Consumption.London: Routledge. Fiske, John (1989), Understanding Popular Culture. Boston: Unwin Hyman. Glaser, BarneyG. and Anselm L. Strauss(1967), TheDiscovery of GroundedTheory.Chicago:Aldine. Grossnickle, Joshua and Oliver Raskin (2000), The Handbookof Online MarketingResearch:Knowing YourCustomerUsing the Net. New York:McGraw-Hill. Gunn, Angela (2000), "Net Ethics: Should You Say Who You Are Online?" Yahoo!InternetLife, 6 (December), 78. Hagel, John and ArthurG. Armstrong(1997), Net Gain: Expanding Markets Through Virtual Communities. Boston: Harvard Business School Press. An Hakken, David (1999), Cyborgs@Cyberspace? Ethnographer Looks to the Future. New York:Routledge. Hamman, Robin (1996), "Cyborgasms: Cybersex Amongst Multiple-Selves and Cyborgs in the Narrow-BandwidthSpace of America Online Chat Rooms," master's dissertation, Department of Sociology, University of Essex, (accessed December 1997) [available at http://www.socio.demon.co.uk/ Cyborgasms.html]. Hammersley, Martyn and Paul Atkinson (1995), Ethnography: Principles in Practice, 2d ed. New York:Routledge. Hirschman,Elizabeth (1986), "Humanistic Inquiryin Marketing Research: Philosophy, Method, and Criteria;" Journal of Marketing Research, 23 (August), 237-49. American Holt, Douglas B. (1998), "Does Cultural CapitalStructure Journalof ConsumerResearch,25 (June), 1-5. Consumption?" Jones, Stephen G. (1995), "UnderstandingCommunity in the InformationAge," in Cybersociety: Computer-Mediated Communication and Community,Stephen G. Jones, ed. Thousand Oaks, CA: Sage Publications, 10-35. ed. (1999), Doing InternetResearch: Critical Issues and -, Methods for Examining the Net. Thousand Oaks, CA: Sage Publications. A Jorgensen, Danny L. (1989), Participant-Observation: Methodology for Human Studies, Vol. 15, Applied Research Methods Series. Newbury Park,CA: Sage Publications.

71
King, Storm A. (1996), "ResearchingInternetCommunities:Proposed Ethical Guidelines for the Reportingof Results," Information Society, 12 (2), 119-27. Kohli, Ajay K. and BernardJ. Jaworski(1990), "MarketOrientation: The Construct, Research Propositions, and Managerial Journalof Marketing,54 (April), 1-18. Implications," Kopytoff, Igor (1986), "The CulturalBiography of Things: Commodification as Process,"in The Social Life of Things: Commodities in Cultural Perspective, Arjun Appadurai,ed. Cambridge, UK: CambridgeUniversityPress, 64-91. Kozinets, RobertV. (1998), "On Netnography:Initial Reflections on Consumer Research Investigations of Cyberculture," in Advances in ConsumerResearch,Vol. 25, JosephAlba and Wesley Hutchinson, ed. Provo, UT: Association for Consumer Research, 366-71. (1999), "E-TribalizedMarketing?The Strategic Implications of VirtualCommunitiesof Consumption," EuropeanManagement Journal, 17 (3), 252-64. Levy, Sidney J. (1959), "Symbols for Sale," Harvard Business Review, 37 (July/August),117-24. Consumer Mythology: A Structural (1981), "Interpreting Approach to Consumer Behavior,"Journal of Marketing, 45 (Summer),49-61. Lincoln, Yvonna S. and Egon G. Guba (1985), Naturalistic Inquiry.Beverly Hills, CA: Sage Publications. McCracken,Grant(1988), The Long Interview.Beverly Hills, CA: Sage Publications. (1990), Cultureand Consumption.Indianapolis,IN: Indiana UniversityPress. Mead, George Herbert(1938), The Philosophy of the Act, Charles W. Morris,ed. Chicago: Universityof Chicago Press. Miller, Daniel and Don Slater (2000), The Internet: An Ethnographic Approach.Oxford, UK: Berg. ComMuniz,AlbertM., Jr.,andThomasC. O'Guinn(2001), "Brand Journalof Consumer Research,27 (March),412-32. munity," Paccagnella, Luciano (1997), "Getting the Seats of Your Pants Researchon VirtualCommuDirty: Strategiesfor Ethnographic nities," Journal of Computer-Mediated Communications, 3 (June) (accessed September 1997) [available at http://www. ascusc.org/jcmc/vol3/issue l/paccagnella.html]. Pendergrast,Mark (1999), UncommonGrounds: The History of Coffee and How It TransformedOur World.New York: Basic Books. Reid, Brian (1995), "USENET Readership Report for March 1995," (accessed December 2000) [available at http://www. Consent in the Study of Reid, Elizabeth(1996), "Informed tlsoft.com/arbitron/mar95/mar95_readership.html]. On-line Communities: A Reflection on the Effects of ComputerMediatedSocial Research," Society, 12 (2), 119-27. Information Rheingold, Howard (1993), The Virtual Community: Homesteading on the Electronic Frontier. Reading, MA: Addison Wesley. Richards,T. and L. Richards(1994), "Using Computersin QualitativeAnalysis,"in TheHandbookof QualitativeResearch,Norman Denzin and Yvonna Lincoln, eds. Thousand Oaks, CA: Sage Publications,445-62. Heart into Schultz, Howardand Dori JonesYang(1999), Pour Your It: How StarbucksBuilt a CompanyOne Cup at a Time. New York:Hyperion. F. Sharf, Barbara (1999), "BeyondNetiquette:The Ethics of Doing Naturalistic Discourse Research on the Internet,"in Doing Internet Research: Critical Issues and Methodsfor Examining the Net, Steve Jones, ed. ThousandOaks, CA: Sage Publications, 243-56. Sherry,John F., Jr. (1991), "PostmodernAlternatives:The Interin pretiveTurnin ConsumerResearch:" Handbookof Consumer Research, Harold H. Kassarjianand Thomas Robertson, eds. Englewood Cliffs, NJ: PrenticeHall, 548-91.

This content downloaded from 137.43.71.99 on Tue, 30 Sep 2014 11:21:55 AM All use subject to JSTOR Terms and Conditions

72

FEBRUARY 2002 OF JOURNAL MARKETING RESEARCH,
Van Maanen,John (1988), Tales of the Field: On WritingEthnography. Chicago: Universityof Chicago Press. Visgaitis, Gary (1996), "On Line's Elusive Face: Researchers Find Net Audience A Moving Target,"USA Today, (October 9), DI. von Hippel, Eric (1986), "LeadUsers: A Source of Novel Product Concepts,"ManagementScience, 32 (July), 791-805. (1988), TheSources of Innovation.New York:Oxford University Press. Wallendorf, Melanie and Russell W. Belk (1989), "Assessing Trustworthinessin NaturalisticConsumer Research,"in Interpretive ConsumerResearch,ElizabethC. Hirschman,ed. Provo, UT: Association for ConsumerResearch,69-84. White, Erin (1999), "'Chatting'a Singer Up the Pop Charts:How Music MarketersUsed the Web to Generate Buzz Before an Album Debuted,"The WallStreetJournal, (October5), B 1, B4. Wittgenstein, Ludwig (1968), Philosophical Investigations, G.E.M. Anscombe, trans. New York:Macmillan.

(1995), "BottomlessCup, Plug-InDrug:A Telethnography of Coffee," VisualAnthropology,7 (4), 351-70. Solomon, Michael R. and HenryAssael (1987), "The Forestor the Trees?A GestaltApproach SymbolicConsumption," Marketto in ing and Semiotics:New Directionsin the Studyof Signsfor Sale, JeanUmiker-Sebeok, New York:Moutonde Gruyter, ed. 189-217. of Spiggle, Susan (1994), "Analysisand Interpretation Qualitative Data in ConsumerResearch," Journalof ConsumerResearch,21 (December), 491-503. Sudweeks, F. and Sheizaf Rafaeli (1995), "How Do YouGet a Hundred Strangers to Agree? Computer-Mediated Communication and Collaboration," Computer in Networkingand Scholarshipin the 21st Century University, Teresa M. Harrisonand Timothy Stephen, ed. New York:SUNY Press, 115-36. Thompson, Craig J. (1997), "Interpreting Consumers: A Hermeneutical Framework for Deriving Marketing Insights from the Texts of Consumers'ConsumptionStories,"Journalof MarketingResearch, 34 (November),438-55. Turkle,Sherry(1995), Life on the Screen: Identityin the Age of the Internet.New York:Simon & Schuster.

This content downloaded from 137.43.71.99 on Tue, 30 Sep 2014 11:21:55 AM All use subject to JSTOR Terms and Conditions