Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.
Eric Yu, Alexei Lapouchnian,and Stephanie DengFaculty of Information andDepartment of Computer ScienceUniversity of Toront...
Motivation IS today function in dynamic, unpredictable world Uncertainties, changing requirements and business andtechno...
Presentation Overview Motivation Example Domain Introduction: BI Adoption Applying social and process modeling techniqu...
The Solution Lifecycle4RequirementsAnalysisDesign/Implementation/ReconfigurationSolutionDeploymentMonitoring/Assessment/Di...
Facing Uncertainties/Change5RequirementsAnalysisDesign/Implementation/ReconfigurationSolutionDeploymentMonitoring/Diagnosi...
Typical Path in BI AdoptionRCIS 2013, 29.05.20136HandlingPUsTraditionalBISelf-ServiceBITailoring toCUsLeveragingSUs12345*n...
Typical Path in BI AdoptionRCIS 2013, 29.05.20137HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs1235*not shown in th...
Typical Path in BI AdoptionRCIS 2013, 29.05.20138HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs1235*not shown in th...
Can Modeling Help? How?RCIS 2013, 29.05.20139ReportinteractivityDDCUBI TeamDStandardReportsAd hocresultDDDDDDDDDAd hocauth...
1: Traditional BIRCIS 2013, 29.05.201310HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs1235*not shown in the present...
1: Traditional BIRCIS 2013, 29.05.201311HandlingPUsTraditionalBISelf-ServiceBITailoring toCUsLeveragingSUs12345*not shown ...
The Case for ModelingRCIS 2013, 29.05.201312HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs12345*not shown in the pr...
RCIS 2013, 29.05.2013 Using the i* social modeling notation to capture the scenario131: Traditional BI: Modeling
RCIS 2013, 29.05.2013 Using the i* social modeling notation to capture the scenario14LEGENDTaskResourceGoalSoftgoalHelpHu...
RCIS 2013, 29.05.2013 Analyzing the model. With i*, we are able to capture: Actors, their functional and non-functional ...
RCIS 2013, 29.05.2013 Analyzing the model. With i*, we are able to capture: Actors, their functional and non-functional ...
RCIS 2013, 29.05.2013 Analyzing the model. With i*, we are able to capture: Actors, their functional and non-functional ...
RCIS 2013, 29.05.2013 Analyzing the model. With i*, we are NOT able to capture: Dynamics of the scenarios – i* model is ...
Process Models: Modeling Temporaland Iterative Aspects Social models Can capture motivations and driving forces behind c...
BusinessMonitoringSell ProductsOnlineAnalyze &InterpretDecideManagerialActionPersonalizeCreate(Query)Execute AnalyzeGenera...
BusinessMonitoringSell ProductsOnlineAnalyze &InterpretDecideManagerialActionPersonalizeCreate(Query)Execute AnalyzeGenera...
BusinessMonitoringSell ProductsOnlineAnalyze &InterpretDecideManagerialActionPersonalizeCreate(Query)Execute AnalyzeGenera...
BusinessMonitoringSell ProductsOnlineAnalyze &InterpretDecideManagerialActionPersonalizeCreate(Query)Execute AnalyzeGenera...
BusinessMonitoringSell ProductsOnlineAnalyze &InterpretDecideManagerialActionPersonalizeCreate(Query)Execute AnalyzeGenera...
A Hierarchical View ofDesign Processes Adaptation loops reveal special relationships amongprocesses Higher-level process...
26
27
28Controller/Designer procs vs. controlled/designed procs• Controller/Designer (C/D) procs – executed less frequently• Dec...
2: Moving to Self-Service BIRCIS 2013, 29.05.201329HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs1235*not shown in ...
2: Moving to Self-Service BIRCIS 2013, 29.05.201330HandlingPUsTraditionalBISelf-ServiceBITailoring toCUsLeveragingSUs12345...
2: Moving to Self-Service BIRCIS 2013, 29.05.201331HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs12345*not shown in...
3: Tailoring to CUsRCIS 2013, 29.05.201332HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs1235*not shown in the prese...
3: Tailoring to CUsRCIS 2013, 29.05.201333HandlingPUsTraditionalBISelf-ServiceBITailoring toCUsLeveragingSUs12345*not show...
34 With i*, we are able to capture: Objectives from the p.o.v. of CUs w.r.t. the tool and assumptions abouttheir capabil...
35CUBI TeamDStandardReportsAd hocresultDD DDDAccess pre-built standardreportingProvide standardreportingDWAd hoc analysisb...
36CUBI TeamDStandardReportsAd hocresultDD DDDAccess pre-built standardreportingProvide standardreportingDWAd hoc analysisb...
3: Tailoring to CUsRCIS 2013, 29.05.201337HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs12345*not shown in the pres...
Conclusions Based on our experience with the case study The need for modeling ongoing adaptation and change in the faceo...
Future Work Further case studies. We are aiming at proposing a modeling and analysisapproach that addresses the challeng...
RCIS 2013, 29.05.201340alexei@cs.toronto.eduQuestions?
Bibliography [1] W. Eckerson, “Business-driven BI using new technologies to fosterself-service access to insights,” TechT...
Upcoming SlideShare
Loading in …5
×

Adapting to Uncertain and Evolving Requirements: The Case of Business-Driven Business Intelligence - RCIS 2013 yu-lapouchnian-deng

388 views

Published on

What happened to "business-driven" BI? Why did the initial attempts not work? How did BI vendors respond? How did user organizations and IT departments adapt? What kinds of modeling can help understand the evolving situation and suggest next steps?

"Adapting to Uncertain and Evolving Requirements: The Case of Business-Driven Business Intelligence" Presentation at the RCIS 2013 conference. Eric Yu, Alexei Lapouchnian, and Stephanie Deng.

Published in: Technology, Business
  • Be the first to comment

  • Be the first to like this

Adapting to Uncertain and Evolving Requirements: The Case of Business-Driven Business Intelligence - RCIS 2013 yu-lapouchnian-deng

  1. 1. Eric Yu, Alexei Lapouchnian,and Stephanie DengFaculty of Information andDepartment of Computer ScienceUniversity of TorontoMay 29, 2013Adapting to Uncertain andEvolving RequirementsThe case of Business-Driven Business Intelligence
  2. 2. Motivation IS today function in dynamic, unpredictable world Uncertainties, changing requirements and business andtechnological landscapes. Ongoing change and adaptation Design/Try/Monitor/Evaluate/Redesign Need new methods for doing analysis/design Most current methods assume stable/predictable setting Need to cope with ongoing change Agility trend – frequent releases Our Objectives Motivate the need for new modeling techniques for such systems Can existing modeling techniques be applied in these dynamicenvironments? In this paper, we concentrate on applying social and process modelsin a socio-technical context2
  3. 3. Presentation Overview Motivation Example Domain Introduction: BI Adoption Applying social and process modeling techniques tothe evolving the BI Adoption scenario Analysis of applicability of the existing notations Discussion Reflections on the use of the existing techniques Wish list for future notations Future WorkRCIS 2013, 29.05.20133
  4. 4. The Solution Lifecycle4RequirementsAnalysisDesign/Implementation/ReconfigurationSolutionDeploymentMonitoring/Assessment/DiagnosisEvolution/AdaptationNeedsNew/ Changed/RefinedRequirementsNewSolutionRequirementsSpecificationData/Observation
  5. 5. Facing Uncertainties/Change5RequirementsAnalysisDesign/Implementation/ReconfigurationSolutionDeploymentMonitoring/DiagnosisEvolution/AdaptationNeedsNew/ Changed/RefinedRequirementsNewSolutionRequirementsSpecificationData/ObservationUncertainty/ChangeUncertainty
  6. 6. Typical Path in BI AdoptionRCIS 2013, 29.05.20136HandlingPUsTraditionalBISelf-ServiceBITailoring toCUsLeveragingSUs12345*not shown in the presentation*…(Further iterationsmay follow)*
  7. 7. Typical Path in BI AdoptionRCIS 2013, 29.05.20137HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs1235*not shown in the presentation*…(Further iterationsmay follow)Business Intelligence is abouttransforming raw data intomeaningful and useful businessinformation• Historical, predictive, current viewsof business dataLeveragingSUs4*
  8. 8. Typical Path in BI AdoptionRCIS 2013, 29.05.20138HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs1235*not shown in the presentation*…(Further iterationsmay follow)Business Intelligence is abouttransforming raw data intomeaningful and useful businessinformation• Historical, predictive, current viewsof business dataContext: Dynamic business environmentFocus: Reporting – bringing BI to BUsExample DomainIterative BI solution adoption path• Common sequence of socio-technicalsolution implementations – evolutionof BI solutions in enterprises• Sources: industry reports [1,2], contactsin the industryLeveragingSUs4*
  9. 9. Can Modeling Help? How?RCIS 2013, 29.05.20139ReportinteractivityDDCUBI TeamDStandardReportsAd hocresultDDDDDDDDDAd hocauthoringtoolAd hoc requestbe fulfilled Ad hocauthoring toolbe providedDDPerform adhoc BI queryExecute BIqueryDDBeimplementedDDSUTimely[ad hoc report]Ad hocreportsDDDDAccess pre-built standardreportingAd hoc analysisbe conductedMonitorbusinessRighttimeFulfilled bySUProvide standardreportingDWMeet businessinformation needsFulfill ad hocrequestInteractivityHelpLearn toolQuick to learn[tool]HelpEmpoweredHelpAnalyticalaptitudeOn-demandtool featureDDHelpTimely[analysis]HelpHelpAd hoc analysisbe supportedProvide ad hocauthoring toolDDBelonging[BI Team]DBI toolexpertiseHelpGovernance[data]DGather inputRecognition[SU]HelpDDHelpFollowenterprisestandardHelpFormulate BIqueryOn-demandtool featureDDTechnicalskillsAccuraterequirementsUserbehaviorSense andanalyze userneedsHelpEnterprisestandard befollowedDToolinputDTool training beprovidedStd. reportrequirementsUserbehaviorDD Evolving Socio-Technical Solutions Process Models: processes, lifecycles, change Social and Goal Models: functional/non-functional objectives,actors, relationships
  10. 10. 1: Traditional BIRCIS 2013, 29.05.201310HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs1235*not shown in the presentation*…(Further iterationsmay follow)Typical Initial BI solution withinmany BI adopters:• BU – standard reports• BI Team – repository ofstandard reports anddevelopment of ad hoc reportsLeveragingSUs4*
  11. 11. 1: Traditional BIRCIS 2013, 29.05.201311HandlingPUsTraditionalBISelf-ServiceBITailoring toCUsLeveragingSUs12345*not shown in the presentation*…(Further iterationsmay follow)Typical Initial BI solution withinmany BI adopters:• BU – standard reports• BI Team – repository ofstandard reports anddevelopment of ad hoc reportsWhat actually happened:• Business domain volatility → changinganalytical needs of BUs• Gap b/w capabilities of standard reports andBU needs grows → # of ad hoc reportrequests increase → bottleneck
  12. 12. The Case for ModelingRCIS 2013, 29.05.201312HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs12345*not shown in the presentation*…(Further iterationsmay follow)LeveragingSUs4*Given what happened,• Could modeling have anticipated the result?• Can modeling be used to help understand whathappened?• Can modeling help us find a new solution?
  13. 13. RCIS 2013, 29.05.2013 Using the i* social modeling notation to capture the scenario131: Traditional BI: Modeling
  14. 14. RCIS 2013, 29.05.2013 Using the i* social modeling notation to capture the scenario14LEGENDTaskResourceGoalSoftgoalHelpHurtMeans-ends linkDecomposition linkContribution linkActorboundaryDDependency link?DeniedUnknown1: Traditional BI: Modeling
  15. 15. RCIS 2013, 29.05.2013 Analyzing the model. With i*, we are able to capture: Actors, their functional and non-functional goals, and dependencies151: Traditional BI: Modeling
  16. 16. RCIS 2013, 29.05.2013 Analyzing the model. With i*, we are able to capture: Actors, their functional and non-functional goals, and dependencies Unmet goals, the driving forces for change161: Traditional BI: Modeling
  17. 17. RCIS 2013, 29.05.2013 Analyzing the model. With i*, we are able to capture: Actors, their functional and non-functional goals, and dependencies Unmet goals, the driving forces for change Alternative ways of fulfilling goals – i.e., the possible adaptation paths171: Traditional BI: Modeling
  18. 18. RCIS 2013, 29.05.2013 Analyzing the model. With i*, we are NOT able to capture: Dynamics of the scenarios – i* model is just a snapshot. Speed and rates of change. Frequencies of occurrences (e.g., of dependencies).18i* Modeling• i* could not have anticipated the failure ofSolution 1, can somewhat help understandwhat happened, can help identify futurealternatives.1: Traditional BI: Modeling
  19. 19. Process Models: Modeling Temporaland Iterative Aspects Social models Can capture motivations and driving forces behind change(+social aspects) We use BPMN (with extensions) to: Look at multiple layers of change in dynamic environments Determine if we can Visualize feedback loops Capture the details of (re-)design cycles, analyze their relativefrequencies, duration.RCIS 2013, 29.05.201319
  20. 20. BusinessMonitoringSell ProductsOnlineAnalyze &InterpretDecideManagerialActionPersonalizeCreate(Query)Execute AnalyzeGenerate(Ad hocreport)Adjust?Analyze(ReportingRequirement)Design Develop TestDeploy(Standardreport)DataWarehousingDrill DownCContinuedD= minutesOperationalProcessAdhocReportingStandardReportingD= seconds ~minutesD= hoursD= daysReport interactivityManagerial ActionStandard reports Ad hoc reportsAd hoc requestReport requirementsOperational dataSCSM SMMSReport requirementsSLegendPoolTaskSequence flowAnnotationSMCSensing flowControl flowMechanism flowStarteventEnd eventStart messageDurationMessage flow DTextEnd messageAssociation
  21. 21. BusinessMonitoringSell ProductsOnlineAnalyze &InterpretDecideManagerialActionPersonalizeCreate(Query)Execute AnalyzeGenerate(Ad hocreport)Adjust?Analyze(ReportingRequirement)Design Develop TestDeploy(Standardreport)DataWarehousingDrill DownCContinuedD= minutesOperationalProcessAdhocReportingStandardReportingD= seconds ~minutesD= hoursD= daysReport interactivityManagerial ActionStandard reports Ad hoc reportsAd hoc requestReport requirementsOperational dataSCSM SMMSReport requirementsSLegendPoolTaskSequence flowAnnotationSMCSensing flowControl flowMechanism flowStarteventEnd eventStart messageDurationMessage flow DTextEnd messageAssociationCapturing Duration
  22. 22. BusinessMonitoringSell ProductsOnlineAnalyze &InterpretDecideManagerialActionPersonalizeCreate(Query)Execute AnalyzeGenerate(Ad hocreport)Adjust?Analyze(ReportingRequirement)Design Develop TestDeploy(Standardreport)DataWarehousingDrill DownCContinuedD= minutesOperationalProcessAdhocReportingStandardReportingD= seconds ~minutesD= hoursD= daysReport interactivityManagerial ActionStandard reports Ad hoc reportsAd hoc requestReport requirementsOperational dataSCSM SMMSReport requirementsSLegendPoolTaskSequence flowAnnotationSMCSensing flowControl flowMechanism flowStarteventEnd eventStart messageDurationMessage flow DTextEnd messageAssociationCapturing Control Flow
  23. 23. BusinessMonitoringSell ProductsOnlineAnalyze &InterpretDecideManagerialActionPersonalizeCreate(Query)Execute AnalyzeGenerate(Ad hocreport)Adjust?Analyze(ReportingRequirement)Design Develop TestDeploy(Standardreport)DataWarehousingDrill DownCContinuedD= minutesOperationalProcessAdhocReportingStandardReportingD= seconds ~minutesD= hoursD= daysReport interactivityManagerial ActionStandard reports Ad hoc reportsAd hoc requestReport requirementsOperational dataSCSM SMMSReport requirementsSIdentifying Sense-Control Pairs
  24. 24. BusinessMonitoringSell ProductsOnlineAnalyze &InterpretDecideManagerialActionPersonalizeCreate(Query)Execute AnalyzeGenerate(Ad hocreport)Adjust?Analyze(ReportingRequirement)Design Develop TestDeploy(Standardreport)DataWarehousingDrill DownCContinuedD= minutesOperationalProcessAdhocReportingStandardReportingD= seconds ~minutesD= hoursD= daysReport interactivityManagerial ActionStandard reports Ad hoc reportsAd hoc requestReport requirementsOperational dataSCSM SMMSReport requirementsSIdentifying Feedback LoopsMonitorAnalyze PlanExecute
  25. 25. A Hierarchical View ofDesign Processes Adaptation loops reveal special relationships amongprocesses Higher-level process – control/design/change process Lower-level processes – target/use/etc. processes Change though Control – constrains the options for the target process. Adaptation. Mechanism – changes the space of options for the target process.Evolution Result – hierarchy of processes reflecting their controlorder. These also help when change cannot be accommodated at runtime(e.g., when we need to design new capabilities)RCIS 2013, 29.05.201325
  26. 26. 26
  27. 27. 27
  28. 28. 28Controller/Designer procs vs. controlled/designed procs• Controller/Designer (C/D) procs – executed less frequently• Decisions in C/D procs – higher-level• Different responsible roles• Different decision criteria• …
  29. 29. 2: Moving to Self-Service BIRCIS 2013, 29.05.201329HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs1235*not shown in the presentation*…(Further iterationsmay follow)In response, vendors came upwith Self-Service BI Tools: shortenad hoc report design cycle• BU – given the self-service tool• BI Team – repository ofstandard reports and tooltrainingLeveragingSUs4*
  30. 30. 2: Moving to Self-Service BIRCIS 2013, 29.05.201330HandlingPUsTraditionalBISelf-ServiceBITailoring toCUsLeveragingSUs12345*not shown in the presentation*…(Further iterationsmay follow)In response: vendors came upwith Self-Service BI Tools, shortenad hoc design cycle• BU – given the self-service tool• BI Team – repository ofstandard reports and tooltrainingWhat actually happened:• Wide-ranging needs of BUs + unrealisticexpectations of BUs’ capabilities → self-service tool too complex or too restrictive.Reverted to Solution 1.
  31. 31. 2: Moving to Self-Service BIRCIS 2013, 29.05.201331HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs12345*not shown in the presentation*…(Further iterationsmay follow)LeveragingSUs4*i* and BPMN Modeling• As we will see in the next iteration, socialmodels could have anticipated the failureand can be used to “diagnose” it.• BPMN models: can’t capture socialaspects, but can analyze shorter reportdesign cycle.
  32. 32. 3: Tailoring to CUsRCIS 2013, 29.05.201332HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs1235*not shown in the presentation*…(Further iterationsmay follow)In response: self-service solutionto Casual Users + customizablestandard reports + ample training• CU – new role (BU with limitedreport creation capabilities).Needs simple and easy-to-learntools.LeveragingSUs4*
  33. 33. 3: Tailoring to CUsRCIS 2013, 29.05.201333HandlingPUsTraditionalBISelf-ServiceBITailoring toCUsLeveragingSUs12345*not shown in the presentation*…(Further iterationsmay follow)In response: self-service solutionto Casual Users + customizablestandard reports + ample training• CU – new role (BU with limitedreport creation capabilities).Needs simple and easy-to-learntools.What actually happened:• Assumption of analytical aptitude fails formany CUs + the tool is difficult even withtraining → limited adoption.
  34. 34. 34 With i*, we are able to capture: Objectives from the p.o.v. of CUs w.r.t. the tool and assumptions abouttheir capabilities w.r.t. the tool and assumptions about their capabilities. NOTE: Physical actors (agents) and logical actors (roles) are distinct in i*. A capability is a property of anagent. Physical agents may fail to possess skills required for playing logical roles.CUBI TeamDStandardReportsAd hocresultDD DDDAccess pre-built standardreportingProvide standardreportingDWAd hoc analysisbe conductedDDAd hocauthoringtoolExecute BIqueryDDBeimplementedD DMonitorbusinessRely on self withauthoring toolMeet businessinformation needsEasy to use[tool]Perform BIqueryEasy to use[tool]DDRighttimeHelpHelp?Learn toolAnalyticalaptitudeQuick to learn[tool]HelpHelp???Zero requestbacklogHelpAd hoc analysisbe supportedProvide ad hocauthoring toolFormulate BIqueryStd. reportrequirementsTool training beprovided3: Tailoring to CUs: Modeling
  35. 35. 35CUBI TeamDStandardReportsAd hocresultDD DDDAccess pre-built standardreportingProvide standardreportingDWAd hoc analysisbe conductedDDAd hocauthoringtoolExecute BIqueryDDBeimplementedD DMonitorbusinessRely on self withauthoring toolMeet businessinformation needsEasy to use[tool]Perform BIqueryEasy to use[tool]DDRighttimeHelpHelp?Learn toolAnalyticalaptitudeQuick to learn[tool]HelpHelp???Zero requestbacklogHelpAd hoc analysisbe supportedProvide ad hocauthoring toolFormulate BIqueryStd. reportrequirementsTool training beprovided3: Tailoring to CUs: Modeling With i*, we are able to capture: Objectives from the p.o.v. of CUs w.r.t. the tool and assumptions abouttheir capabilities. Failures to achieve CUs’ goals and wrong capability assumption.
  36. 36. 36CUBI TeamDStandardReportsAd hocresultDD DDDAccess pre-built standardreportingProvide standardreportingDWAd hoc analysisbe conductedDDAd hocauthoringtoolExecute BIqueryDDBeimplementedD DMonitorbusinessRely on self withauthoring toolMeet businessinformation needsEasy to use[tool]Perform BIqueryEasy to use[tool]DDRighttimeHelpHelp?Learn toolAnalyticalaptitudeQuick to learn[tool]HelpHelp???Zero requestbacklogHelpAd hoc analysisbe supportedProvide ad hocauthoring toolFormulate BIqueryStd. reportrequirementsTool training beprovided3: Tailoring to CUs: Modeling With i*, we are NOT able to capture: The difference in frequencies of Learn Tool and Execute BI Query cannotbe captured. Gradual learning and skill improvement and tool adjustment cannot beeasily represented.
  37. 37. 3: Tailoring to CUsRCIS 2013, 29.05.201337HandlingPUsTraditionalBISelf-ServiceBITailoring toCUs12345*not shown in the presentation*…(Further iterationsmay follow)LeveragingSUs4*i* and BPMN Modeling• i* could have predicted some problems withsolution adoption and can help with postmortem.• BPMN models: can’t capture socialaspects, but can analyze shorter reportdesign cycle.
  38. 38. Conclusions Based on our experience with the case study The need for modeling ongoing adaptation and change in the faceof uncertain/evolving requirements is a fact of life for enterprises. Co-evolution and alignment of the social and technical componentsof solutions is important. Neither the i* social modeling notation nor the (augmented) BPMNnotation is adequate. Aspects of modeling and analysis that are important tosupport: Variability Modeling and Binding, criteria for alternative selection,barriers to adoption/change. Social Modeling – physical vs. logical actors, skills/capabilities,personal goals, incentives. Feedback (failures, changes in context, requirements), multiplelevels of design – changes within/across levels, iterations. Temporal and dynamic aspects – frequencies, duration, etc.38
  39. 39. Future Work Further case studies. We are aiming at proposing a modeling and analysisapproach that addresses the challenges identified in thispaper. Working on the idea of enterprise decision space Requirements-driven, goal-oriented High-variability, explicit capturing of decision criteria Many linked decisions at multiple levels Feedback Informed by BIRCIS 2013, 29.05.201339
  40. 40. RCIS 2013, 29.05.201340alexei@cs.toronto.eduQuestions?
  41. 41. Bibliography [1] W. Eckerson, “Business-driven BI using new technologies to fosterself-service access to insights,” TechTarget, 2012. [2] B. Evelson, “The Forrester wave: self-service business intelligence platforms,Q2, 2012”, Forrester Research, 2012.41

×