Deep Data Analytics for New Product Development

New product failuresDesign failuresPricing failuresMessaging failuresAn NPD processThe heart of the NPD processMarket researchBusiness analyticsSummaryNotesIdeation: What do you do?Sources for ideasTraditional approachesA modern approachBig Data – external and internalText data and text analysisDocuments, corpus, and corporaOrganizing text dataText processingCreating a searchable databaseCall center logs and warranty claims analysisSentiment analysis and opinion miningMarket research: voice of the customer (VOC)Competitive assessment: the role of CEAContextual designMachine learning methodsManaging ideas and predictive analyticsSoftwareSummaryAppendixMatrix decompositionSingular value decomposition (SVD)Spectral and singular value decompositionsDevelop: How do you do it?Product design optimizationConjoint analysis for product optimizationConjoint frameworkConjoint design for new productsA new product design exampleConjoint designSome problems with conjoint analysisOptimal attribute levelsSoftwareKansei engineering for product optimizationStudy designsCombining conjoint and Kansei analysesEarly-stage pricingvan Westendorp price sensitivity meterSummaryAppendix 3.ABrief overview of the chi-square statisticAppendix 3.BBrief overview of correspondence analysisAppendix 3.CVery brief overview of ordinary least squares analysisBrief overview of principal components analysisPrincipal components regression analysisBrief overview of partial least squares analysisTest: Will it work and sell?Discrete choice analysisProduct configuration vs. competitive offeringsDiscrete choice background – high-level viewTest market hands-on analysisLive trial tests with customersMarket segmentationTURF analysisSoftwareSummaryAppendixTURF calculationsLaunch I: What is the marketing mix?Messaging/claims analysisStages of message analysisMessage creationMessage testingPrice finalizationGranger–Gabor analysisPrice segmentationPricing in a social networkPlacing the new productSoftwareSummaryLaunch II: How much will sell?Predicting vs. forecastingForecasting responsibilityTime series and forecasting backgroundData issuesData availabilityTraining and testing data setsForecasting methods based on data availabilityNaive methodsSophisticated forecasting methodsData requirementsForecast error analysisSoftwareSummaryAppendixTime series definitionBackshift and differencing operatorsRandom walk model and naive forecastRandom walk with driftConstant mean modelThe ARIMA family of modelsTrack: Did you succeed?Transactions analysisBusiness intelligence vs. business analyticsBusiness intelligence dashboardsThe limits of business intelligence dashboardsCase studyCase study data sourcesCase study data analysisPredictive modelingNew product forecast error analysisAdditional external data – text once moreSentiment analysis and opinion miningSentiment methodology overviewSoftwareSummaryAppendixDemonstration of linearization using log transformationDemonstration of variance stabilization using log transformationConstant elasticity modelsTotal revenue elasticityEffects tests F-ratiosResources: Making it workThe role and importance of organizational collaborationAnalytical talentTechnology skill setsData scientists, statisticians, and machine learning expertsConstant trainingSoftware issuesDownplaying spreadsheetsOpen source softwareCommercial softwareSQL: A must-know languageOverall software recommendationJupyter/Jupyter LabBibliography
Next >