Evaluating Packaged vs Best of Breed BI Solutions. Wayne Eckerson. Director, TDWI Research

Transcription

1 Build or Buy? Evaluating Packaged vs Best of Breed BI Solutions Wayne Eckerson Director, TDWI Research

2 Agenda Industry Trends Convergence BI Platforms Packaged Analytic Applications Best of Breed

3 Market Evolution Business Intelligence Performance Management Office of CEO/CFO Data Warehousing BI as foundation technology BI for the midmarket 1990s 2000s 2010

4 Convergence Operational Systems Analytical Systems ERP: Analytical systems are an afterthought BI/DW Analyze historical data in batch

5 Convergence Operational Systems Analytical Systems ERP: Analytical systems are a growth area BI/DW: Monitor process performance

6 Convergence Operational Systems Analytical Systems ERP: Embed analytical systems BI/DW: Analytical systems close the loop.

7 Consolidation ERP Players Oracle SAP IBM Microsoft BI Players Hyperion Business Objects Cognos Proclarity MicroStrategy Teradata SAS Institute Hewlett Packard Information Builders Informatica Actuate Others

8 Consolidation ERP Players Oracle/Hyperion SAP/BO IBM/Cognos Microsoft/Proclarity BI+ Players BI Players SAS Institute Teradata Hewlett Packard MicroStrategy Information Builders Informatica Actuate Others

9 The Others Best of Breed? In-memory BI/Visualization QlikTech, Spotfire, Tableau, Advizor Solutions Dashboards Corda, idashboards, Visual Mining, Appfusion, Theoris MPP Databases Aster, Greenplum, Kognitio DW Appliances Netezza, Dataupia, Kickfire, Teradata Columnar Databases Sybase, Vertica, ParAccel, InfoBright Event driven analytics Streambase, SeeWhy, Truviso, Syndera, Truviso Data Integration GoldenGate, Expressor, ab Initio, Pervasive, iway, DataFlux Open Source Pentaho, Jaspersoft, Talend, Apatar, Software-as-a-Service LucidEra, PivotLink, Oco,

10 Build vs Buy Technology platforms Packaged analytic applications Best of breed

11 BI Stack Legacy Technical Team Business Users Applica- Data tions Files, Web Services Extract Clean Model Transform Transfer Load Data Data Warehousing Environment Query Report Analyze Mine Visualize Act Data Integration Environment Reporting & Analysis Environment

12 Evolution of Suites ETL Data Quality Data Profiling Metadata ETL Suite

13 ETL Evolution of Platforms Production Reporting Data Quality Data Profiling End-user Reporting End-user Analysis Metadata Change data Data Capture Movement Database OLAP DashboardsScorecards d EII/ Federation Server Operating System Visualization Source Adapters Text/Externa Data Storage Planning Budgeting Master Data Mgmt Data Integration Environment Data Warehousing Environment Analytics Reporting & Analysis Environment

14 Source App Source System Source App Evolution of Platforms Data Movement ETL Data Data End-user Quality Profiling Reporting Metadata Production Reporting End-user Analysis Database DashboardsScorecards The Big Four Oracle SAP Microsoft IBM Source System Change data EII/ Capture Federation Server Operating Visualization System OLAP Source App Source Adapters Text/Externa Data Storage Planning Budgeting Source System Data Integration Environment Master Data Mgmt Data Warehousing Environment Analytics Reporting & Analysis Environment

15 Where do you obtain your BI tools? Best-of-breed tools from multiple vendors 55% Integrated suite from a single vendor 46% From applications we purchase 22% We will build them 13% From Wayne Eckerson and Cindi Howson, Enterprise Business Intelligence: Strategies and Technologies for Deploying BI on an Enterprise Scale, Based on 594 respondents.

16 Best of Breed vs Platforms Best of Breed Suite/platform Production Reporting OLAP End-user Reporting End-user Analysis -Focused -Innovative -Holistic/unified -Responsive -Value pricing But -Comprehensive enough? -Integrated with 3 rd party tools? -Viable? -All in one -Integrated -High ROI -Single contact -Volume pricing But -Degree of integration? -Degree of functionality? -Vendor focus?

17 Standardization Success Rates To What Degree has Your BI Standardization Been Successful? 38% 33% 20% 1% 8% Very high High Moderate Low Very low From Wayne Eckerson and Cindi Howson, Enterprise Business Intelligence: Strategies and Technologies for Deploying BI on an Enterprise Scale, Based on 594 respondents.

18 Standardization Challenges Individual resistance to change Departmental autonomy 43% 48% High switching costs 28% Training and support costs 26% Lack of executive support 17% Negotiating pricing with vendors 12% From Wayne Eckerson and Cindi Howson, Enterprise Business Intelligence: Strategies and Technologies for Deploying BI on an Enterprise Scale, Based on 594 respondents.

19 Packaged Analytic Applications End-to-end application built on a BI platform that supports the reporting and analytical requirements of a specific business domain. Examples: CRM analytics for consumer retailers Od Order-to-ship reporting and analysis Supplier analysis for discrete manufacturers Workforce planning and scorecards Fraud and money laundering analytics

20 Packaged Analytic Apps - Classic kaged Applicatio on Pac X Source Adapters Adapter Domain specific application Metadata ETL ETL Mappings Database Data Model d Semantic Layer End-user Reporting End-user Analysis Dashboards Reports Closed Loop

21 Packaged Analytic Apps BI Tools Domain specific application Database End-user Reporting End-user Analysis Data Model Semantic Layer Reports

22 Packaged Analytic Apps: SaaS Package ed Applic cation X Source Adapters Adapter Metadata ETL ETL Mappings Database Data Model End-user Reporting Dashboards End-user Analysis Semantic Reports Layer We b Brows ser On premis e Web Service On premis e

23 Packaged Analytic Apps: DaaS Package ed Applic cation X Source Adapters Adapter Metadata ETL ETL Mappings Database Data Model End-user Reporting Dashboards End-user Analysis Semantic Reports Layer We b Brows ser On premis e Web Service On premis e

24 Packaged Analytic Applications Reso ources PACKAGED User feedback Configure application Refine reports Test and deploy CUSTOM Get requirements Scope project Hire developers Design schema Build ETL Document metadata Build reports Get feedback Test, implement Train users Time to Deploy

25 Evaluating Packaged Applications Pros Speed to deployment and ROI Incorporates industry best practices One throat to choke Closed loop processing Cons Silos of applications Duplicate components Lack of configuration flexibility Difficult to upgrade customizations Vendor lock in

26 Customizing Packages Based on 162 respondents. Cost of ownership increases

27 Evaluating Build Your Own Pros Get what you want Pay only for what you build Use existing tools and expertise Differentiate from the competition Cons Cost of maintenance Development skills required Developers must be available May not have broad knowledge of best practices

28 Why Build? Why did you build the application?

29 The Build Option What did you use to build the application?

30 Keys to Success Buying Packages Use model-driven packages Configure, don t customize Look for software-as-a-service as a service packages Building Applications Build when it s a competitive differentiator Don t reinvent the wheel Build with industry standard frameworks Document rules in metadata repository for easy maintenance

31 Summary The BI industry is consolidating Specialists vs Platforms vs Apps Pros and cons to each To get the best value: Understand vendor s resources Understand vendor s state of mind Understand vendor s future And know thyself! Understand your organization s culture and capacity To determine whether build or buy really makes sense

32 Contact Wayne Eckerson Director, TDWI Research