Emerald InsightInternationalJournalofPhysical Distribution&LogisticManagementAglobalexplorationof BigDatainthesupplychainRobertGlennRicheyJrTylerR.MorganKristinaLindsey-HallFrankG.AdamsArticleinformation:To cite this document:Cd)RobertGlennRicheyJrTylerR.MorganKristinaLindsey-HallFrankG.Adams,(2016)"AglobalexplorationofBigData inthesupplychain",International Journal ofPhysicalDistribution&Logisticso2Management,Vol.46lss8pp.710-739Permanentlinktothisdocument:0http://dx.doi.org/10.1108/lJPDLM-05-2016-0134-7Downloadedon:29November2016,At:22:37(PT)2References: this document contains references to 5o other documents.Tocopythisdocument:permissions@emeraldinsight.com2Thefulltextofthisdocumenthasbeendownloaded1004times since2016*SUserswhodownloadedthisarticlealsodownloaded:(2016)."Amulti-agentbasedsystemwithbigdataprocessingforenhancedsupplychainagility"0JourmalofEnterpriseInformationManagement,Vol.29lss5pp.706-727http:lidx.doi.org/io.1108/JEIM-06-2015-0050000(2016),"Socialnetworkanalysis in supplychain management research",International JournalofPhysicalDistribution&LogisticsManagement,Vol.46Iss8pp.740-762http:/l3dx.doi.org/10.1108/IJPDLM-05-2015-0122Access to this document was granted through an Emerald subscription provided by emerald-Busrm:313548[]ForAuthors2If youwould liketowriteforthis,or anyother Emeraldpublication,thenpleaseuseourEmeraldforAuthors serviceinformationabouthowtochoosewhichpublicationtowriteforand submissionguidelinesareavailableforall.Pleasevisitwww.emeraldinsight.com/authorsformoreinformation.LIMOCAboutEmeraldwww.emeraldinsight.comEmeraldisa global publisherlinkingresearchand practicetothebenefitofsociety.Thecompanymanagesaportfolioof morethan290 journalsandover2,350booksandbookseriesvolumes,aswell as providingan extensiverangeof onlineproducts and additional customer resources andservices.Emerald isbothCOUNTER4andTRANSFERcompliant.Theorganization isapartneroftheCommitteeonPublicationEthics(COPE)andalsoworkswithPorticoandtheLOCKSSinitiativefordigitalarchivepreservation.*Related content and download information correct at time ofdownload
International Journal of Physical Distribution & Logistics Management A global exploration of Big Data in the supply chain Robert Glenn Richey Jr Tyler R. Morgan Kristina Lindsey-Hall Frank G. Adams Article information: To cite this document: Robert Glenn Richey Jr Tyler R. Morgan Kristina Lindsey-Hall Frank G. Adams , (2016),"A global exploration of Big Data in the supply chain", International Journal of Physical Distribution & Logistics Management, Vol. 46 Iss 8 pp. 710 - 739 Permanent link to this document: http://dx.doi.org/10.1108/IJPDLM-05-2016-0134 Downloaded on: 29 November 2016, At: 22:37 (PT) References: this document contains references to 50 other documents. To copy this document: permissions@emeraldinsight.com The fulltext of this document has been downloaded 1004 times since 2016* Users who downloaded this article also downloaded: (2016),"A multi-agent based system with big data processing for enhanced supply chain agility", Journal of Enterprise Information Management, Vol. 29 Iss 5 pp. 706-727 http://dx.doi.org/10.1108/ JEIM-06-2015-0050 (2016),"Social network analysis in supply chain management research", International Journal of Physical Distribution & Logistics Management, Vol. 46 Iss 8 pp. 740-762 http:// dx.doi.org/10.1108/IJPDLM-05-2015-0122 Access to this document was granted through an Emerald subscription provided by emeraldsrm:313548 [] For Authors If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service information about how to choose which publication to write for and submission guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information. About Emerald www.emeraldinsight.com Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online products and additional customer resources and services. Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation. *Related content and download information correct at time of download. Downloaded by Huazhong University of Science and Technology At 22:37 29 November 2016 (PT)
The current issue and full text archive of this journal is available on Emerald Insight at:www.emeraldinsight.com/0960-0035.htmIJPDLMA global exploration of Big Data46,8in the supply chainRobert Glenn RicheyJrDepartment of Systems and Technology,Auburn University710Auburn,Alabama, USATylerR.MorganReceived 13 May 2016(d)Revised 18 Jume 2016Department of SupplyChain and InformationSystemsAcoxpted 18 Jume 2016100lowaStateUniversity,Ames,Iowa,USAKristina Lindsev-HallDepartmentofMarketing,UniversityofAlabama,Tuscaloosa,Alabama,USA,andFrank G. AdamsDepartmentofMarketing,Mississippi StateUniversityStarkville,Mississippi,USAAbstractPurpose-Journals inbusinesslogistics,operationsmanagement,supplychainmanagement,andbusiness strategyhave initiated ongoing calls for Big Data researchand its impact on research andpractice. Currently, no extant research has defined the concept fully. The purpose of this paper is todevelop an industry grounded definition of Big Data by canvassing supply chain managers across sixooaeonations.The supply chain setting defines BigData as inclusive offour dimensions:volume,velocityvariety,and veracity.The study further extracts multiple concepts that are important to thefuture ofsupply chain relationship strategy and performance.These outcomes provide a starting point andextend a call for theoretically grounded and paradigm-breaking research on managing business-to-business relationships in the age of Big Data.Design/methodology/approach -A native categories qualitative method commonlyemployed in sociology allows each executive respondent to provide rich, specific data.Thisapproach reduces interviewer bias while examining 27 companies across six industrialized andindustrializing nations.This is the first study in supply chain management and logistics (SCMLs) touse the native category approach.Findings -This study defines Big Data by developing four supporting dimensions that inform andground future SCMLsresearch;details ten key success factors/issues;and discusses extensiveopportunitiesforfutureresearchResearch limitations/implications -This study provides a central grounding of the term,dimensions,and issues related toBig Data in supply chainresearch.Practical implications-Supplychain managers are provided with a peer-specific definition andunified dimensions of Big Data.The authors detail key success factors for strategic consideration.Finally,this studynotes differencesin relational priorities concerming these success factors acrossdifferentmarkets,andpoints tofuture complexity in managing supply chain and logistics relationshipsOriginality/value-There is currently no central grounding of the term, dimensions, andEmeraldissues related to Big Data in supply chain research. Forthe first time,the authors addresssubjects related to how supply chain partners employ Big Dataacross the supply chain, uncover BigData's potential to influence supply chain performance, and detail the obstacles to developing Bigntematiopal lourmalofPhvsicalData'spotentialIn addition,the study introducesthenative category qualitativeinterviewapproachtoDistribution & LogisticsSCMLs researchersManagemerVoL46 No.8.2016Keywords Big Data, Governance,Effectiveness, Relationships, Efficiency,Integration, GlobalPp.710-73TransparencyOEmerald Group Publishing Limited0960-0039Paper type Research paperDO 101108/JFDEM05-20160134
A global exploration of Big Data in the supply chain Robert Glenn Richey Jr Department of Systems and Technology, Auburn University, Auburn, Alabama, USA Tyler R. Morgan Department of Supply Chain and Information Systems, Iowa State University, Ames, Iowa, USA Kristina Lindsey-Hall Department of Marketing, University of Alabama, Tuscaloosa, Alabama, USA, and Frank G. Adams Department of Marketing, Mississippi State University, Starkville, Mississippi, USA Abstract Purpose – Journals in business logistics, operations management, supply chain management, and business strategy have initiated ongoing calls for Big Data research and its impact on research and practice. Currently, no extant research has defined the concept fully. The purpose of this paper is to develop an industry grounded definition of Big Data by canvassing supply chain managers across six nations. The supply chain setting defines Big Data as inclusive of four dimensions: volume, velocity, variety, and veracity. The study further extracts multiple concepts that are important to the future of supply chain relationship strategy and performance. These outcomes provide a starting point and extend a call for theoretically grounded and paradigm-breaking research on managing business-tobusiness relationships in the age of Big Data. Design/methodology/approach – A native categories qualitative method commonly employed in sociology allows each executive respondent to provide rich, specific data. This approach reduces interviewer bias while examining 27 companies across six industrialized and industrializing nations. This is the first study in supply chain management and logistics (SCMLs) to use the native category approach. Findings – This study defines Big Data by developing four supporting dimensions that inform and ground future SCMLs research; details ten key success factors/issues; and discusses extensive opportunities for future research. Research limitations/implications – This study provides a central grounding of the term, dimensions, and issues related to Big Data in supply chain research. Practical implications – Supply chain managers are provided with a peer-specific definition and unified dimensions of Big Data. The authors detail key success factors for strategic consideration. Finally, this study notes differences in relational priorities concerning these success factors across different markets, and points to future complexity in managing supply chain and logistics relationships. Originality/value – There is currently no central grounding of the term, dimensions, and issues related to Big Data in supply chain research. For the first time, the authors address subjects related to how supply chain partners employ Big Data across the supply chain, uncover Big Data’s potential to influence supply chain performance, and detail the obstacles to developing Big Data’s potential. In addition, the study introduces the native category qualitative interview approach to SCMLs researchers. Keywords Big Data, Governance, Effectiveness, Relationships, Efficiency, Integration, Global, Transparency Paper type Research paper International Journal of Physical Distribution & Logistics Management Vol. 46 No. 8, 2016 pp. 710-739 © Emerald Group Publishing Limited 0960-0035 DOI 10.1108/IJPDLM-05-2016-0134 Received 13 May 2016 Revised 18 June 2016 Accepted 18 June 2016 The current issue and full text archive of this journal is available on Emerald Insight at: www.emeraldinsight.com/0960-0035.htm 710 IJPDLM 46,8 Downloaded by Huazhong University of Science and Technology At 22:37 29 November 2016 (PT)
IntroductionGlobal"Big Data"is industry's hottest buzzword (see Waller and Fawcett,2013).Consultantsexplorationproclaim that Big Data will revolutionize industry and decision making.Currentof BigDatainvestments suggest Big Data's importance:International Data Corporation“predictsthat the market for Big Data will reach $16.1billion in 2014,grow six times faster thantheoverall IT marketij”Still,some practitioners fearthatBigData'svalue is711exaggerated and viewitas littlemore than an extension of forecasting ortraditionalmarketingresearch.Unfortunately,asisthecasewith manyhottopicsand buzzwordssupply chain research with useful implications is lacking."[Tjhere is very little published()9102management scholarship that tackles the challenges of using such tool; or better yet,thatexploresthepromiseandopportunitiesfornewtheoriesandpracticesthatBigDatamightbringabout"(Georgeetal,2014,p.321).Infact,thereisalmostnoconsistencyindefining Big Data, identifying its purpose, or establishing its role in supply chainmanagement. Thanks to a grant from the Council of Supply Chain ManagementProfessionals,we seek answers to questions about whether supply chain partnersemployBigData in supply chain strategy,thetheoretical originsof BigData'spotential2toinfluencesupplychainperformance,andtheobstaclestodeveloping itspotentialGiven the scarcity of research on Big Data as a supply chain construct, it is wellsuited to qualitativeexploration.The strength ofqualitativeresearch isthat itprovidesinsights regarding difficult-to-perceive issues in supply chain management andlogistics(SCMLs)(Manganetal,2004;Naslund,2002).Estimates of theuseofqualitativemethodsinSCMresearchvarywidely.GolicicandDavis(2012)notethat"[...J the reported frequency of studies using qualitative methods varies a great deal;puethisislikelyduetothedifferentjournalsincluded inthedifferentreviews.Onereviewreports qualitativemethods in as fewas1o percent of the articles,and onefinds5l percent of the studies they examined to use qualitative methods (this particularstudyincludedanyjournalinwhich'supplychainmanagementwasa subjectofthearticle)"(p.729).However,Fawcett and Waller (2011) estimate that only 10-20 percentof published articles in SCML embracequalitative methods.In general,qualitativeresearch is considered effective in addressing the“how"questions regarding processesbeing studied (Pratt, 2009).uoeapeounThe questions posed here stem froma need to better understand Big Data's role insupply chains (George et al.,2014),and this study describes the qualitative method usedto examine those questions. The research then explores subjects'comments todetermine how closely their views correspond to extant-if ill-defined -definitions ofBig Data.Further,weexaminehow supplychainmanagers useBig Data toexposeBigData'skey success factors and enablers in supply chain management.Thekey success factors and enablers described here suggest that Big Data, as asupply chain construct, should be characterized as structured and unstructuredrelationship-basedinformation unique to itsholder because of the information'svolume, velocity,variety,and veracity.Further, while such a definition might implythat Big Data is nomorethanan undifferentiated resource, theresponses explored herealso suggest that the collectiveabilities to assembleand leverage Big Data constitute afirm's dynamic capability (DC).ResearchquestionsThe discussion of“what"Big Data“is"remains incomplete.Industry provides no completeandspecificdefinition.Popularliteraturediscussesvolume,velocity,variety,andveracitybutgiveslimiteddefinitionsofeach.SupplychainmanagersremainconfusedaboutBig
Introduction “Big Data” is industry’s hottest buzzword (see Waller and Fawcett, 2013). Consultants proclaim that Big Data will revolutionize industry and decision making. Current investments suggest Big Data’s importance: International Data Corporation “predicts that the market for Big Data will reach $16.1 billion in 2014, grow six times faster than the overall IT market[1].” Still, some practitioners fear that Big Data’s value is exaggerated and view it as little more than an extension of forecasting or traditional marketing research. Unfortunately, as is the case with many hot topics and buzzwords, supply chain research with useful implications is lacking. “[T]here is very little published management scholarship that tackles the challenges of using such tool; or better yet, [that] explores the promise and opportunities for new theories and practices that Big Data might bring about” (George et al., 2014, p. 321). In fact, there is almost no consistency in defining Big Data, identifying its purpose, or establishing its role in supply chain management. Thanks to a grant from the Council of Supply Chain Management Professionals, we seek answers to questions about whether supply chain partners employ Big Data in supply chain strategy, the theoretical origins of Big Data’s potential to influence supply chain performance, and the obstacles to developing its potential. Given the scarcity of research on Big Data as a supply chain construct, it is well suited to qualitative exploration. The strength of qualitative research is that it provides insights regarding difficult-to-perceive issues in supply chain management and logistics (SCMLs) (Mangan et al., 2004; Näslund, 2002). Estimates of the use of qualitative methods in SCM research vary widely. Golicic and Davis (2012) note that “[.] the reported frequency of studies using qualitative methods varies a great deal; this is likely due to the different journals included in the different reviews. One review reports qualitative methods in as few as 10 percent of the articles, and one finds 51 percent of the studies they examined to use qualitative methods (this particular study included any journal in which ‘supply chain management’ was a subject of the article)” (p. 729). However, Fawcett and Waller (2011) estimate that only 10-20 percent of published articles in SCML embrace qualitative methods. In general, qualitative research is considered effective in addressing the “how” questions regarding processes being studied (Pratt, 2009). The questions posed here stem from a need to better understand Big Data’s role in supply chains (George et al., 2014), and this study describes the qualitative method used to examine those questions. The research then explores subjects’ comments to determine how closely their views correspond to extant – if ill-defined – definitions of Big Data. Further, we examine how supply chain managers use Big Data to expose Big Data’s key success factors and enablers in supply chain management. The key success factors and enablers described here suggest that Big Data, as a supply chain construct, should be characterized as structured and unstructured relationship-based information unique to its holder because of the information’s volume, velocity, variety, and veracity. Further, while such a definition might imply that Big Data is no more than an undifferentiated resource, the responses explored here also suggest that the collective abilities to assemble and leverage Big Data constitute a firm’s dynamic capability (DC). Research questions The discussion of“what” Big Data “is” remains incomplete. Industry provides no complete and specific definition. Popular literature discusses volume, velocity, variety, and veracity, but gives limited definitions of each. Supply chain managers remain confused about Big 711 Global exploration of Big Data Downloaded by Huazhong University of Science and Technology At 22:37 29 November 2016 (PT)
JPDLMData's creation and sourcing (Richey et al,2015).This study treats Big Data as bothstrategicandoperational,soourfirststepwastodevelopanunderstandingof what46,8supplychain managers thinkBig Data is from theirperspectiveThisdefinitionmustbeconfirmed beforeaccuratefutureresearchcanbeperformed (Richey et al,2015)A lack of clear, easily understood terms has long challenged supply chainresearchers.For example,the word “agility"is used to mean“adaptability,712“flexibility"“slack"and“improvization"(Gligor and Holcomb, 2012, Golicic andSebastiano, 2011).From a relationships perspective, it is nearly impossible formanagers tocollaboratewhen theylack a consistent interpretation of common terms.aeaCouncil of SupplyChainManagementProfessional's (CSCMP)listofacronyms,as wellas the use of Incoterms,are prominent examples of managing and simplifyingterminology.Overall, it is important to reach consensus regarding a definition of BigData" in SCML relationships. We should derive this definition through a managementperspectiveratherthanfromacommercialorganizationperspectivebecausethelatter'sgoal is to enticefirms topurchase theirproducts and services inhopes ofunlocking BigData's potential performance.Followingabasic conceptualization (orperhaps re-conceptualization)of BigDatawe seek to develop an initial understanding of Big Data's future in supplychainstrategy and operations.Fromabest practicesperspective,managers often ask aboutthekey success factors to making relationship-based processes work (Fawcett andMagnan, 2002; Richey et al.,2010).At the onset of new supply chain initiatives,managers assess existing capabilities and options.So,the next task is uncovering BigData-related, performance-based, key success factors.Categorizing, assessing quality, and identifying Big Data's impact is very new tomanagement in general (George et al,2014).Managers hope supply chain relationshipswill create synergy that allowsfirms to increase revenue and/or reduce costs beyondwhat a single firm can accomplish. Thus, while it is important to understandopportunities, it is equally important to understand potentially negative consequencesthat may reduce or eliminate firm and joint-firm performance.Though essential tosupply chain management, software developers and managers often overlook thesedynamics because they are embedded in technology departments with limitedboundary-spanningopportunities.Our studyaddressestheseresearchquestions:RQ1.HowdoesBigData impact supplychains?RQ2.Whatarethekey successfactors forBigData implementation in SCMIL?RQ3.Whattheorybestexplainsthepotentiallong-termvalueBigData canbringtosupply chain partnerships?This study uses a lived experience qualitative interview technique to address thesevery broad questions.These open-ended questions are intentionally general andprovide limited structure to allow respondents to answer at length, thus reducinginterviewer bias. We discuss this method briefly in the following section.Next, wediscuss our industry-basedresultsand conclusionsfromaglobalperspective,includingChina, Germany, India, South Korea, Turkey, and the USA.ResearchmethodWe used a native categoryapproach in our research methodology.Often used insociology and anthropology research, the native categories approach avoids interviewer
Data’s creation and sourcing (Richey et al., 2015). This study treats Big Data as both strategic and operational, so our first step was to develop an understanding of what supply chain managers think Big Data is from their perspective. This definition must be confirmed before accurate future research can be performed (Richey et al., 2015). A lack of clear, easily understood terms has long challenged supply chain researchers. For example, the word “agility” is used to mean “adaptability,” “flexibility,” “slack,” and “improvization” (Gligor and Holcomb, 2012; Golicic and Sebastiano, 2011). From a relationships perspective, it is nearly impossible for managers to collaborate when they lack a consistent interpretation of common terms. Council of Supply Chain Management Professional’s (CSCMP) list of acronyms, as well as the use of Incoterms, are prominent examples of managing and simplifying terminology. Overall, it is important to reach consensus regarding a definition of “Big Data” in SCML relationships. We should derive this definition through a management perspective rather than from a commercial organization perspective because the latter’s goal is to entice firms to purchase their products and services in hopes of unlocking Big Data’s potential performance. Following a basic conceptualization (or perhaps re-conceptualization) of Big Data, we seek to develop an initial understanding of Big Data’s future in supply chain strategy and operations. From a best practices perspective, managers often ask about the key success factors to making relationship-based processes work (Fawcett and Magnan, 2002; Richey et al., 2010). At the onset of new supply chain initiatives, managers assess existing capabilities and options. So, the next task is uncovering Big Data-related, performance-based, key success factors. Categorizing, assessing quality, and identifying Big Data’s impact is very new to management in general (George et al., 2014). Managers hope supply chain relationships will create synergy that allows firms to increase revenue and/or reduce costs beyond what a single firm can accomplish. Thus, while it is important to understand opportunities, it is equally important to understand potentially negative consequences that may reduce or eliminate firm and joint-firm performance. Though essential to supply chain management, software developers and managers often overlook these dynamics because they are embedded in technology departments with limited boundary-spanning opportunities. Our study addresses these research questions: RQ1. How does Big Data impact supply chains? RQ2. What are the key success factors for Big Data implementation in SCML? RQ3. What theory best explains the potential long-term value Big Data can bring to supply chain partnerships? This study uses a lived experience qualitative interview technique to address these very broad questions. These open-ended questions are intentionally general and provide limited structure to allow respondents to answer at length, thus reducing interviewer bias. We discuss this method briefly in the following section. Next, we discuss our industry-based results and conclusions from a global perspective, including China, Germany, India, South Korea, Turkey, and the USA. Research method We used a native category approach in our research methodology. Often used in sociology and anthropology research, the native categories approach avoids interviewer 712 IJPDLM 46,8 Downloaded by Huazhong University of Science and Technology At 22:37 29 November 2016 (PT)
historical bias by sequentially and increasingly removing the interviewer from directionGlobalof the interview process (Harris,2000).The approach is designed to force respondentsexplorationintoopen-endeddiscussionswiththeinterviewer,centeringfuturequestioningonthekeyofBigDatacomments made by the respondent.This method of qualitative research provesespeciallyvaluable and appropriate when the phenomena studied arenotpreciselyunderstood, as well as when the researcher needs to draw perspective from subjects713experiencingthephenomena (Corbin andStrauss,2008).Thisallowsrelevantfactorstoemerge from the data. In line with naturalistic views (Lincoln and Guba, 1985),this studyemployspurposive sampling (Kuzel,1992)and reputational case selection(Goetz andLeCompte, 1984), seeking out subjects likely to have a rich understanding of ourphenomenon of interest, but withvaried perspectives.Purposive sampling is often usedinqualitativeresearchwhenthesubjectsselectedpossessexperiencecongruentwiththeresearchobjective(Guestetal.2006).UsingtheCSCMIPmembershiprosterandknownindustryparticipants,weidentified managers from several countrieswho areknown forusing information technology across supply chain relationships.Wesoughttodevelopabreadthofunderstandingacrossseveral countries.Assessing academic research, popular press articles, and industry participants'locations,we identified nine countries:Brazil, China, Germany,India, South AfricaSouth Korea, Turkey, the UK, and the USA. Based on existing qualitative interviewconventions,we setagoal ofacquiring a minimum ofthreeinterviews in each countrySampling and data collectionSubjects in Brazil, Germany,South Africa, and theUK werereluctant to discusscompanystrategy,soweaugmentedoursamplebyusinga comparablecaseapproach(Goetzand LeCompte,1984)and solicited memberstovolunteer contact information ofother executives qualified to participate in the study.The referral approach waspositive in Germany (six respondents), but resulted in only single completed interviewsinBrazil and SouthAfricaand no responses from the UK.The final sample included three interviews each from China, India, South Korea, andTurkey,six interviews in Germany,and nine in the USA[2] Becausethe low response madeitimpossibletoassesssaturation,ourdataanalysisexcludestranscriptsfromBrazil,SouthAfrica,andtheUKTableIprovideseachrespondentfirm'snationalityindustry,employeesize,and supplychainclassification.Tohelpinsuretrustworthinessofthedatacollection(see Table I, we restricted the firmographic data heavily to ensure confidentiality.QualitativeinterviewquestionsandanalysisThenative category approach allowed us to dig deeply into managers'unobservableand subjectiveperceptions,while using a naive line of inquiry in order to avoid leadingquestions thatcould impact interviewer expectations and answers (Harris,2000).Thisapproach also allowed us to ask very basic questions in a dynamic way.Table IIpresents an example of the native questionnaire tool, and the specific questions areprovided in the Appendix.Most interviews iasted between 30 and 70 minutes, depending on response detail andcomplexity.Wedidnotrecordfiveoftheinterviewsbecauseofcompanypolicies.Whenrecordingwasnotavailable,weuseddualnotetakersandrepetitiontechniquestoensureaccuracy.WeconductedallinterviewsinEnglishwiththeexceptionoftheSouthKoreanones.English impeded interviewcompletion withSouthKoreanrespondents,sothesewereconductedinKoreanandthentranslatedtoEnglishforanalysis
historical bias by sequentially and increasingly removing the interviewer from direction of the interview process (Harris, 2000). The approach is designed to force respondents into open-ended discussions with the interviewer, centering future questioning on the key comments made by the respondent. This method of qualitative research proves especially valuable and appropriate when the phenomena studied are not precisely understood, as well as when the researcher needs to draw perspective from subjects experiencing the phenomena (Corbin and Strauss, 2008). This allows relevant factors to emerge from the data. In line with naturalistic views (Lincoln and Guba, 1985), this study employs purposive sampling (Kuzel, 1992) and reputational case selection (Goetz and LeCompte, 1984), seeking out subjects likely to have a rich understanding of our phenomenon of interest, but with varied perspectives. Purposive sampling is often used in qualitative research when the subjects selected possess experience congruent with the research objective (Guest et al., 2006). Using the CSCMP membership roster and known industry participants, we identified managers from several countries who are known for using information technology across supply chain relationships. We sought to develop a breadth of understanding across several countries. Assessing academic research, popular press articles, and industry participants’ locations, we identified nine countries: Brazil, China, Germany, India, South Africa, South Korea, Turkey, the UK, and the USA. Based on existing qualitative interview conventions, we set a goal of acquiring a minimum of three interviews in each country. Sampling and data collection Subjects in Brazil, Germany, South Africa, and the UK were reluctant to discuss company strategy, so we augmented our sample by using a comparable case approach (Goetz and LeCompte, 1984) and solicited members to volunteer contact information of other executives qualified to participate in the study. The referral approach was positive in Germany (six respondents), but resulted in only single completed interviews in Brazil and South Africa and no responses from the UK. The final sample included three interviews each from China, India, South Korea, and Turkey, six interviews in Germany, and nine in the USA[2]. Because the low response made it impossible to assess saturation, our data analysis excludes transcripts from Brazil, South Africa, and the UK. Table I provides each respondent firm’s nationality, industry, employee size, and supply chain classification. To help insure trustworthiness of the data collection (see Table I), we restricted the firmographic data heavily to ensure confidentiality. Qualitative interview questions and analysis The native category approach allowed us to dig deeply into managers’ unobservable and subjective perceptions, while using a naive line of inquiry in order to avoid leading questions that could impact interviewer expectations and answers (Harris, 2000). This approach also allowed us to ask very basic questions in a dynamic way. Table II presents an example of the native questionnaire tool, and the specific questions are provided in the Appendix. Most interviews lasted between 30 and 70 minutes, depending on response detail and complexity. We did not record five of the interviews because of company policies. When recording was not available, we used dual note takers and repetition techniques to ensure accuracy. We conducted all interviews in English with the exception of the South Korean ones. English impeded interview completion with South Korean respondents, so these were conducted in Korean and then translated to English for analysis. 713 Global exploration of Big Data Downloaded by Huazhong University of Science and Technology At 22:37 29 November 2016 (PT)
(pamuos)IJPDLMe insassSeeeeeandaeeBeeesse46,8eineroeae"saouoe000'-00001SE714o000'000sunoorosenp(odk)nsnpupaerooasenepuenodeueeaoaoanooeiopeaK.nsnpusissaoSGpresseanSnenuetSeeaoaseeSnpeueoareinssaneneaeesraoBo=0SS0色ooua na00000-000001000'09-000'001000T-000g000'T-000'S000'T-000'SvsnsuoonoonooisiueaeopadKnspuaded(aa)Table I.apoRespondents33S55
USA Industrial nations BRIC/MINTS Code Industry (type) Firm employee count Job title, supply chain classification Code Industry (type) Firm employee count Job title, supply chain classification Code Industry (type) Firm employee count Job title, supply chain classification U1 Paper products 100,000-50,000 Project integration manager, manufacturing G1 Automotive and aerospace 50,000-10,000 President and CEO, 3PL/ 4PL/ consulting C1 Kitchen and bath products 10,000-5,000 Chief information officer, manufacturing U2 Automotive 5,000-1,000 Purchasing and logistics manager, manufacturing G2 Automotive industry o100 Vice president, supplier/ distributer C2 Automotive 50,000- 10,000 IT supply chain manager, manufacturing U3 Automotive 100,000-50,000 System engineer, manufacturing G3 Information technology and system integration 5,000-1,000 Head of departure management, 3PL/4PL/ Consulting C3 Electronics and infrastructure 50,000- 10,000 Asset controller, manufacturing U4 3PL and publishing 5,000-1,000 Senior director of global business development, manufacturing G4 Sporting and athletic goods 100,000-50,000 Director of supply chain management, manufacturing I1 Technology engineering and infrastructure W300,000 American material manager, manufacturing U5 Healthcare 5,000-1,000 Purchasing director, retailing G5 Land transport, ocean and air freight, and contract logistics 100,000-50,000 Senior VP of IT, transportation I2 Automotive 50,000- 10,000 Deputy general manager, manufacturing (continued ) Table I. Respondents 714 IJPDLM 46,8 Downloaded by Huazhong University of Science and Technology At 22:37 29 November 2016 (PT)
Globaleddns o peeSeeaeerBearpreoessexplorationsoosoeaoepereeamnof Big Data三000T-000's000T-000's000T-000'S715-00000000aunn100aoeeaAnsnpa)pFZL33Sreeednsoioee KddnseuoeiouSuinsnoonad qorridtAS0a00001-00000suoneuenspl000'T-000S000'T-000'S001>une00ose(ok)nspupoope presaaoruoneuioyuooasaapoepedospue poodaspoos902soaooenSnoenueSeanemuoenorednosoe.o0SanosSugeary perangBb色品roa00001-000'00000'T-000S000000000vSnooA^Ansnp(aa)apoTable I.89n5B
USA Industrial nations BRIC/MINTS Code Industry (type) Firm employee count Job title, supply chain classification Code Industry (type) Firm employee count Job title, supply chain classification Code Industry (type) Firm employee count Job title, supply chain classification U6 Retail W150,000 Lead for corporate initiatives and transportation, retailing G6 Cosmeticscustomer packaged goods 50,000-10,000 Chief supply chain officer, manufacturing I3 Water and environment management 5,000-1,000 Head of field management, manufacturer U7 Dairy 50,000-10,000 Director supply chain, manufacturing K1 Research and development o100 Information service provider, 3PL/ 4PL/ consulting T1 Automotive 5,000-1,000 Group chairman and CEO, manufacturing U8 Energy 5,000-1,000 Strategic sourcing, manufacturing K2 Food and beverage 5,000-1,000 Manager, retailing T2 Automotive 5,000-1,000 Head of supply chain and operations, manufacturing and retailing U9 Retail W150,000 Director of transportation technologies, retailing K3 Transportation/ information technology 5,000-1,000 Senior fellow, 3PL/4PL/ consulting T3 Food and beverage 50,000- 10,000 Multichannel operations manager, retailing Table I. 715 Global exploration of Big Data Downloaded by Huazhong University of Science and Technology At 22:37 29 November 2016 (PT)
IJPDLMWhat are theFirst obstacleSecond obstacleThird obstacleFourth obstacleFifthobstacleobstacles towards46,8using Big Data incategory issuecategory issuecategory issuecategory issuecategory issueyour Supply Chainrelationships?Importance:,Importance:Importance:Importance:Importance:Rank:Rank:Rank:Rank.RankWhyisarobstacle?1//716How does thisobstacle impactyour relationshipperformance?()9100Who in therelationship isXimpactedthemosby this obstacle?Is there a way foryour supply chain toovercome thisobstacle?"Youhave"You have"You haveYou have"Youhave2discussed [firstdiscussed [seconddiscussed [thirddiscussed [fourthdiscussed [fifthLE:issue]. Whenissue].Whenissue]. Whenissuel.Whenissue]. Whenthinking about thethinking about thethinking about thethinking about thethinking about theTable II.future of yourfutureofyourfuture of yourfutureofyourfuture ofyourIVNative categorybusiness, are therebusiness, are therebusiness, are therebusiness, are therebusiness, are thereABflow exampleother things thatother things thatother things thatother things thatother things thatnconsideryuconsidervouconsidergyou consider?"you consider?Following Wallendorf and Belk (1989),we designed the data collection and analysis toLpueadhereto three dimensions of trustworthiness: integrity,confirmability,and triangulationaeaeoToassureintegrity-meaningguardingagainstfabrications-respondentswereassuredconfidentiality; we reference all respondents only by codes, and we identify firms only bysize,industry,and nationality.Confirmability means consistently applied techniques thatenablemultipleresearcherstoarriveseparatelyatsimilarconclusions.Weattainedthisdimension by employing three separate researchers to examine all interviews individuallyusing open coding,with specific attention tocommonalitiesanddifferenceswithinandbetweennationalities.Finally,data triangulationmeansmultipleresearchers'findings canreconciletoa commonmeaning.Weachieved triangulation throughmultiplemeetings of thecoding researchers to comparefindings.Afterwe identified themes,each coding researcherre-examined the data to determine if any ofthefindingsdid not correspond to comments inthe data.Additionally,a separate memberof the research team,who was not part of thecoding process,examined thefindings tofurther strengthen the analyses'confirmabilityResultsFollowing coding completion, the research team examined the results for concepts thatsurfaced from thedata relevant to the three research questions motivating this study.Thedescription of results begins with the definitions that emerged,followed bykeysuccess factors and enablers.Finally,we discuss applicabletheory as it relates tothevalue Big Data can bring to supply chains.Whatsupplychainmanagers call“BigData"[Big] Data contain a wide variety [of information] I see so many databases made withoutpurpose or definition. Everyone needs their own definition of Big Data in order to use BigData in a productive way (K3)
Following Wallendorf and Belk (1989), we designed the data collection and analysis to adhere to three dimensions of trustworthiness: integrity, confirmability, and triangulation. To assure integrity – meaning guarding against fabrications – respondents were assured confidentiality; we reference all respondents only by codes, and we identify firms only by size, industry, and nationality. Confirmability means consistently applied techniques that enable multiple researchers to arrive separately at similar conclusions. We attained this dimension by employing three separate researchers to examine all interviews individually using open coding, with specific attention to commonalities and differences within and between nationalities. Finally, data triangulation means multiple researchers’ findings can reconcile to a common meaning. We achieved triangulation through multiple meetings of the coding researchers to compare findings. After we identified themes, each coding researcher re-examined the data to determine if any of the findings did not correspond to comments in the data. Additionally, a separate member of the research team, who was not part of the coding process, examined the findings to further strengthen the analyses’ confirmability. Results Following coding completion, the research team examined the results for concepts that surfaced from the data relevant to the three research questions motivating this study. The description of results begins with the definitions that emerged, followed by key success factors and enablers. Finally, we discuss applicable theory as it relates to the value Big Data can bring to supply chains. What supply chain managers call “Big Data” [Big] Data contain a wide variety [of information]. I see so many databases made without purpose or definition. Everyone needs their own definition of Big Data in order to use Big Data in a productive way (K3). Table II. Native category flow example 716 IJPDLM 46,8 Downloaded by Huazhong University of Science and Technology At 22:37 29 November 2016 (PT)
Managers around the world careabout Big Data.Yet, we quicklylearned that there isGlobalno consensus among supply chain managersregarding Big Data's definition.Theexplorationrespondentsinthisstudydiscussedacontinuumofdata--fromunstructured"tweet"ofBigDatabased data to internal logistics and operations indicators - suggesting that thediscussion of Big Data maybe confounded by a lack of clearly differentiated andinterpreted terms. One respondent touched on Big Data's capacity to continually717examinethecompetitivelandscape:I am not familiar with that wording [Big Data] My understanding for what you are after isthat you want to understand how we are dealing with the oneversion of the truth, forexample, in terms of data information.What weare using in our group is Bl, which isBusiness Intelligence, and the different systems in our landscape, in our supply chainorganization are feeding the data information and should therefore result into one version ofthe truth (G4).Academia and consultancyprovide terms that defineBig Data's context,butthis maydiffertremendouslyfrompractitioners.This inconsistency mayberesponsible inpartforthecommunicationproblem(s)amongtop management,logisticsmanagers,datascientists,consultants,and softwaremanufacturerswhoare only nowattemptingtointegrateBigDataintosupplychainprocesses.Thefollowinganalysisprovidesanextracteddefinition ofBigData for supplychainmanagersand fleshesoutthemeaningof the four dimensions being discussed in popular literature. While managers maydescribe Big Data differently,more than one respondent discussed using Big Data toseize and configureresources to copewith changing trends:Ithink it's having access to and making sense of your industry data and taking advantage of it.as best you can, using it to drive decisions and to see trends. It's grasping at allthe information,putting it into some meaningful models, and using it to drive decision-making (U9).These types of responses that detail using Big Data to increase visibility ofdemand and improve responsiveness wereprevalentthroughoutthedata collectionprocess across all countries.Yet managers seemed to have differing understandingsof what Big Datais in formTo build aplatform for futureacademic scrutiny,wehavesynthesizedwhatwediscoveredamongthesesixcountriesandrespondentsCentral to the issue's complexity is that each respondent supply chain managerdefines Big Data according to his or her own needs from operational and strategicmindsets.Forinstance:To me the Big Data, more like the orders/details in supply chain, including like how manyshipments each shipment, the route for each shipment, the weight, the CBMs for eachshipment, things like that. So it's the very details of our daily operations in supply chain (Cl)This is a decidedly operational view of Big Data, focussed heavily on resource visibilityWhilethe corporation's internal data are large,growing,and thus certainlyBig,thecomment ishardly consistent with what SAS callsBig Data,namely,“theexponentialgrowth and availability of data, both structured and unstructured[3]" Other managerstake the concept to the 20,o00-foot view, perhaps beyond the SAS definition:Compiling all data in huge databases available across countries and business unit records (GI)This supply chain manager considers Big Data to betruly multi-level, multi-firm,andmulti-industry data across all locations and countries.This broad viewencompasseswhat most distressessupplychainmanagers-howdowemakethis data useful?andhowdowemakeviablewhatwecannot define strategicallyand operationally?McAfee
Managers around the world care about Big Data. Yet, we quickly learned that there is no consensus among supply chain managers regarding Big Data’s definition. The respondents in this study discussed a continuum of data – from unstructured “tweet”- based data to internal logistics and operations indicators – suggesting that the discussion of Big Data may be confounded by a lack of clearly differentiated and interpreted terms. One respondent touched on Big Data’s capacity to continually examine the competitive landscape: I am not familiar with that wording [Big Data]. My understanding for what you are after is that you want to understand how we are dealing with the one version of the truth, for example, in terms of data information. What we are using in our group is BI, which is Business Intelligence, and the different systems in our landscape, in our supply chain organization are feeding the data information and should therefore result into one version of the truth (G4). Academia and consultancy provide terms that define Big Data’s context, but this may differ tremendously from practitioners. This inconsistency may be responsible in part for the communication problem(s) among top management, logistics managers, data scientists, consultants, and software manufacturers who are only now attempting to integrate Big Data into supply chain processes. The following analysis provides an extracted definition of Big Data for supply chain managers and fleshes out the meaning of the four dimensions being discussed in popular literature. While managers may describe Big Data differently, more than one respondent discussed using Big Data to seize and configure resources to cope with changing trends: I think it’s having access to and making sense of your industry data and taking advantage of it, as best you can, using it to drive decisions and to see trends. It’s grasping at all the information, putting it into some meaningful models, and using it to drive decision-making (U9). These types of responses that detail using Big Data to increase visibility of demand and improve responsiveness were prevalent throughout the data collection process across all countries. Yet managers seemed to have differing understandings of what Big Data is in form. To build a platform for future academic scrutiny, we have synthesized what we discovered among these six countries and respondents. Central to the issue’s complexity is that each respondent supply chain manager defines Big Data according to his or her own needs from operational and strategic mindsets. For instance: To me the Big Data, more like the orders/details in supply chain, including like how many shipments each shipment, the route for each shipment, the weight, the CBMs for each shipment, things like that. So it’s the very details of our daily operations in supply chain (C1). This is a decidedly operational view of Big Data, focussed heavily on resource visibility. While the corporation’s internal data are large, growing, and thus certainly Big, the comment is hardly consistent with what SAS calls Big Data, namely, “the exponential growth and availability of data, both structured and unstructured[3].” Other managers take the concept to the 20,000-foot view, perhaps beyond the SAS definition: Compiling all data in huge databases available across countries and business unit records (G1). This supply chain manager considers Big Data to be truly multi-level, multi-firm, and multi-industry data across all locations and countries. This broad view encompasses what most distresses supply chain managers – how do we make this data useful? and how do we make viable what we cannot define strategically and operationally? McAfee 717 Global exploration of Big Data Downloaded by Huazhong University of Science and Technology At 22:37 29 November 2016 (PT)
JPDLMandBrynjolfsson (2012)pioneered thediscussion ofBigDatabydefining itaccordingto three specific aspects: the volume of information produced, the velocity at which it is46,8created,andthevarietyofformsittakes.SASexpandedthisdefinitiontoincludetheterms/dimensions ofcomplexityand variability.All of thesedimensions were evidentin our research but we found that Big Data could best be described through fourdimensions:volume,velocity,variety,and veracity.718Volumewas an issue ofcomprehensiveconcern.Respondents in all countries notedthe immeasurableamountof availabledataand theneedtomanage itto helptiemarketinformation todecision processes:aeaHow can you get sensible [...J information out of the Big Data volumes or enough data to beable to support [and], efficiently be able to analyze the strategies or improve the company (G3)Many of the respondents understand that thevolume of data being handled makessupplychainmanagement'sfutureverydifferentfromthepast.Despitedisagreementsabout the data's value,general agreement exists concerning the huge amount of dataavailable.How to handle the data's growth is a different concern.All respondents also noted the velocity of data growth as a defining characteristicthat makes the era of Big Data different from past data management projects.Considering themountinguseofRFIDtags,scannerdata,WMS,ERP,etc,thatprovidereal-time information, we can see a mountingrateofdata speed growth (Cakici et al,2011;Zhongetal,2015).Likemanyweinterviewed,respondentU2notedconcernsabouthis/her supply chain being able tokeeppacewithBig Data growth.Nearly everyrespondent noted the increased velocity of the data being produced. As discuss later,speed is a defining aspectof BigDataas supplychain managersconsidervelocitytobebothanobstacleandanopportunity.The huge variety of data sources is more cryptic, but is openly discussed by supplychain managers:[...Jthe combination of mainly bringing all different data areas together, and by having moreand more diverseportfolios or un-harmonized, Imean just that you bring things undermarket, likeapps,you bring things under market likeweb servicesor web solutions.Youbringthiskindofstuff,letmesaymoreandmoreintotheoperationsprocesses.Then,vouhave to takecare of theprep work that are connected so to say to“backroom or backbone,ofnotgettingreallyspreadout withthiskind of data.Andforsureitis alsothe increaseof thedata volume,while you haveall kinds of connections, which we havefrom outside into ourcompany,and fromtheinsideoutofourcompany.So,maybethatwhatweseeas[...|all theinteractions, which we haveto establish and to maintain and also then the collections betweenall our systems (G5).Data today comes in countlessformats from numerous locations.Most supplychainmanagers areaccustomedto structured data (accounting,operational numbers,etc.)This is largely qualitative data translated into numeric form for traditionaldatabases. Most of it is operational in context and can be within company or acrossintegrated partnerships. Unstructured data, or data that is not pre-defined forsoftware analysis,presents themorechallenging and less-used data source.Forms ofunstructured data include text documents, e-mail, audio, financial transactionstweets,andevenvideo.Supplychainmanagersarebeginningtolearnhowtotapthevast varieties of availabledata, such as addressing logistics-related customer servicecomplaints via Twitter (Bhattacharjya et al, 2016). One Turkish respondent (T2)brings the first three dimensions together in his/her definition of Big Data bycommenting that Big Data is“fast,""big,"and“diverse
and Brynjolfsson (2012) pioneered the discussion of Big Data by defining it according to three specific aspects: the volume of information produced, the velocity at which it is created, and the variety of forms it takes. SAS expanded this definition to include the terms/dimensions of complexity and variability. All of these dimensions were evident in our research but we found that Big Data could best be described through four dimensions: volume, velocity, variety, and veracity. Volume was an issue of comprehensive concern. Respondents in all countries noted the immeasurable amount of available data and the need to manage it to help tie market information to decision processes: How can you get sensible [.] information out of the Big Data volumes or enough data to be able to support [and], efficiently be able to analyze the strategies or improve the company (G3). Many of the respondents understand that the volume of data being handled makes supply chain management’s future very different from the past. Despite disagreements about the data’s value, general agreement exists concerning the huge amount of data available. How to handle the data’s growth is a different concern. All respondents also noted the velocity of data growth as a defining characteristic that makes the era of Big Data different from past data management projects. Considering the mounting use of RFID tags, scanner data, WMS, ERP, etc., that provide real-time information, we can see a mounting rate of data speed growth (Çakıcı et al., 2011; Zhong et al., 2015). Like many we interviewed, respondent U2 noted concerns about his/her supply chain being able to keep pace with Big Data growth. Nearly every respondent noted the increased velocity of the data being produced. As discuss later, speed is a defining aspect of Big Data as supply chain managers consider velocity to be both an obstacle and an opportunity. The huge variety of data sources is more cryptic, but is openly discussed by supply chain managers: [.] the combination of mainly bringing all different data areas together, and by having more and more diverse portfolios or un-harmonized, I mean just that you bring things under market, like apps, you bring things under market like web services or web solutions. You bring this kind of stuff, let me say more and more into the operations processes. Then, you have to take care of the prep work that are connected so to say to “backroom or backbone,” of not getting really spread out with this kind of data. And for sure it is also the increase of the data volume, while you have all kinds of connections, which we have from outside into our company, and from the inside out of our company. So, maybe that what we see as [.] all the interactions, which we have to establish and to maintain and also then the collections between all our systems (G5). Data today comes in countless formats from numerous locations. Most supply chain managers are accustomed to structured data (accounting, operational numbers, etc.). This is largely qualitative data translated into numeric form for traditional databases. Most of it is operational in context and can be within company or across integrated partnerships. Unstructured data, or data that is not pre-defined for software analysis, presents the more challenging and less-used data source. Forms of unstructured data include text documents, e-mail, audio, financial transactions, tweets, and even video. Supply chain managers are beginning to learn how to tap the vast varieties of available data, such as addressing logistics-related customer service complaints via Twitter (Bhattacharjya et al., 2016). One Turkish respondent (T2) brings the first three dimensions together in his/her definition of Big Data by commenting that Big Data is “fast,” “big,” and “diverse.” 718 IJPDLM 46,8 Downloaded by Huazhong University of Science and Technology At 22:37 29 November 2016 (PT)