Slide Heading Understanding the Business Intelligence Framework Michael J Scarbrough 09 November 2015 Introductions • Michael J Scarbrough, CISA, CRISC, MBA – VP-Senior Audit Manager @ BMO (Bank of Montreal) Financial Group – Corporate Audit Division – Data analysis – Data modeling of controls for 100% effectiveness testing – TDWI Executive Committee – Committee to develop control framework methodology for BI environments in 2005-2009 – Expert at auditing BI environments (since 2004) Agenda Defining Business Intelligence Business Intelligence Framework Business Intelligence Risks-Exposures-Controls Slide Heading Metadata Types Metadata Sources Business Intelligence: History • BI formally coined in 1998 • Dynamic & evolving at a rapid pace • BI Common Body of Knowledge maintained by TDWI. – The Data Warehouse Institute: “TDWI is dedicated to educating business and information technology professionals about the strategies, techniques, and tools required to successfully design, build, and maintain business intelligence and data warehousing solutions. It also fosters the advancement of business intelligence and data warehousing research and contributes to knowledge transfer and professional development of its Members.” Credits: The Data Warehouse Institute 4 Business Intelligence: Definition • “…a set of concepts and methodologies to improve decision making in business through use of facts and fact-based systems.” --Howard Dresner, The Gartner Group • “An architecture & collection of integrated operational as well as decision support applications & databases that provide the business community easy access to business data. --Larissa T. Moss & Shaku Arte, BI Roadmap • “…processes, technologies, & tools needed to turn data into information, information into knowledge, and knowledge into plans that drive profitable business actions.” --David Loshin, BI: The Savvy Manager’s Guide Credits: The Data Warehouse Institute 5 BI Requirements • Organizational skills & motivation to develop a BI program where results are applied back into the business. • Knowledge & skills to use business analysis to identify/create business information • Collections of quality data & metadata important to the business • Application of analytic tools, techniques, & processes Credits: The Data Warehouse Institute, MJ Scarbrough 6 BI as a Data “Refinery” BI As a Data Refinery Credits: The Data Warehouse Institute 7 TDWI: BI Components Framework TDWI Business Intelligence Framework: April 2004 Business Requirements BI Architecture Data Warehousing/Data Mart Data Resource Administration Data Acquisition, Cleansing, & Integration Development BI & DW Operations Program Management Data Sources Data Stores Information Services Information Delivery Business Analytics Business Applications Business Value Credits: The Data Warehouse Institute, MJ Scarbrough 8 1. 2. 3. 4. 5. 6. 7. Business Layer BI results meet business requirements BI program is business owned & driven BI contributes to strategic positioning BI is a key enabler of business tactics Solutions are aligned with business processes Business metrics are meaningful & valued BI delivers measurable business value Business Value · Business Impact · Strategic Positioning · Tactical Effectiveness · Insight & Innovation · Organization Learning & Growth BI Component Descriptions Business architecture guides business applications. Applications (business uses) are central to creating business value Value is created by meeting business requirements Business Requirements · Respond to business drivers · Achieve business goals · Implement business strategies · Enable & inform business tactics · Optimize business processes · Meet business information needs Development · Data resources · Information services · Business metrics · Dashboards & scorecards · Analytic applications Program Management · Charter & sponsorship · Cohesive projects · Organizational impact · Managed quality · Measures of success 1. 2. 3. 4. 5. 6. 7. Getting from requirements to value demands a commitment to program mgmt & a sound approach to development Sustainable BI requires sound administrative & operations practices **Items are color coordinated w/BI Framework Components tab Infrastructure (Data Store) Layer Data manipulation & derivation standards Identify data gaps Paint the business picture & integrate the technology Engineer the data warehouse for value, revenue, & cost reduction Data integration Data quality Data warehousing is the “engine” that drives business intelligence 1. 2. 3. 4. 5. 6. 1. 2. 3. 4. 5. 6. Administration & Operations Layer The business is connected with the data Decision-makers have access to information Business analytics support business mgmt Information is of high quality Information services are reliable BI services adapt & change with the business BI Architecture · Business architecture · Information/data architecture · Process architecture · Organizational architecture · Technology architecture Business Applications · Performance management · Business analytics · Actionable information · Discovery & forecasting · Decision support Data Resource Administration · Data governance · Data ownership · Data stewardship · Data custodianship · Data security BI & DW Operations · Scheduling · Execution · Support · Maintenance Information Services Layer Query & reporting OLAP (Online Analytical Processing) DSS/EIS (Decision Support System/Executive Information System) Analytic applications Data Mining Information services turns warehousing results into BI products. Credits: The Data Warehouse Institute, MJ Scarbrough 9 Business Layer (Green) • Business value – Benefits anticipated • Business requirements – Reasons to implement BI – Results required (e.g. metrics, information needs) • Development – Project activities that create & deploy BI & DW/DM products • Program Management – Ongoing activities to manage BI program Credits: The Data Warehouse Institute, MJ Scarbrough 10 Business Layer Goals • • • • • • BI delivers business value Results meet business requirements BI program is business owned & driven BI contributes to strategic positioning BI is a key enabler of business tactics Solutions aligned with business processes • Identified and integrated business metrics are value-added Credits: The Data Warehouse Institute, MJ Scarbrough 11 Administration & Operations Layer (Yellow) • BI Architecture – Framework, standards, & conventions that describe BI environment components • Business applications – Business processes & procedures that access/receive information for results • Data Resource Administration – Data governance, including data owner, steward, and custodian • BI & DW/DM Operations – Execution, monitoring, and maintaining quality, availability, and performance Credits: The Data Warehouse Institute, MJ Scarbrough 12 Administration & Operations Layer Goals • • • • • • Business is connected with the data Decision makers have access to information Business analytics support business objectives Information is of high quality Information services & resources are reliable BI services adapt & change with the business Credits: The Data Warehouse Institute, MJ Scarbrough 13 Infrastructure (Data Store) Layer (Blue) • BI Implementation & Infrastructure – Technical components needed to: • Capture data • Turn data into value-added information • Deliver information to the business Credits: The Data Warehouse Institute, MJ Scarbrough 14 Infrastructure Layer Goals • Data manipulation & derivation standards • Identify data gaps • Ensure the business information is integrated with the technology • Engineer the DW/DM for to meet business objectives • Data integration • Data quality • DW/DM acting as a successful technology to support BI Credits: The Data Warehouse Institute, MJ Scarbrough 15 Information Services Layer (Orange) • BI Analytical Components that place information into business function context (i.e CRM, SCM, BPM, etc.). • Would include: – Static Reporting – Dashboards – Scorecards Credits: The Data Warehouse Institute, MJ Scarbrough 16 Information Services Layer Goals • Information Services turn DW/DM results into the following BI products: – Query & reporting – OLAP (Online Analytical Processing) – DSS/EIS (Decision Support System/Executive Information System) – Analytic Applications – Data Mining Credits: The Data Warehouse Institute, MJ Scarbrough Information Services Layer Implementation The Landscape for Analytical Tools Credits: The Data Warehouse Institute, MJ Scarbrough 18 BI Technology Framework CRM Analytics Desktop Email Wireless Analytic Applications BPM (Scorecards & Dashboards) Supply Chain Analytics Operations Analytics Analytic Apps Development Tools, Packages, Templates B2E Portal (Intranet) Infrastructure Web Collaboration Email, Groupware, Workflow Data Access & Analysis Query, Reporting, OLAP, Mining, Forecasting Text Analysis Text Search & Text Mining Data Warehouse / Data Marts Content Management Data Integration B2B & B2C Portals (Internet/Extranet) Storage, Servers, Databases, Metadata, Administration & Management, Networking TDWI Business Intelligence Technology Framework: April 2004 Modeling, Mapping, Cleansing, ETL Data Resources Operational Systems, Documents, Images, External Data, Audio/Visual Credits: The Data Warehouse Institute, MJ Scarbrough 19 BI Organizational Framework TDWI Business Intelligence Organizational Framework: April 2004 BI Program Sponsorship Program Management Data Governance BI Projects Database Development ETL Development Project Management Data Integration & Cleansing Data Access, Analysis & Mining Business Metrics Usage System & Database Administration Process Execution & Monitoring Training & Support Data Mart Development Metadata Management Business Requirements Definition BI Operations Source Data Analysis Business Rules Specification Integration Design BI Application Development Architecture Specification Quality Management Credits: The Data Warehouse Institute, MJ Scarbrough 20 BI Corporate Governance Structure 1. Establish Corporate BI Steward • Accountability for overall information management 2. Establish a Corporate Domain Responsible for BI/IM P&P 3. Establish Corporate Custodianship for Enforcement/Exception Approval • Programs, Projects, Operations 21 BI/IM Governance Risks 1. Business decisions executed on inaccurate data and/or models maintained in data stores. 2. Increased costs and overhead associated with line of business vs. consortium, enterprise purchases of BI-related software and hardware. 3. Incompatibility of data elements and systems when information is shared between lines of business and networks, impeding the ability to successfully cross sell products across lines of business. 4. Non-alignment of innovation and/or research and development within lines of business, resulting in unsupported, non-compliant technologies and inefficient use of resources. 5. Data integrity compromised due to a lack of uniform data and/or its associated dimensions across lines of business and/or system platforms. 22 BI/IM Governance Risks 6. Impediment of executive information and customer strategies due to disparate, non-uniform policies, procedures, and guidelines among lines of business. 7. Incomplete or inaccurate data resulting from inadequate monitoring of interfacing data feeds originating from the systems of record. 8. Exposure of private and protected customer information maintained outside of business line systems of record. 9. Increased latency resulting from inefficient and/or undocumented design and structure of a non-uniform BI environment. 10. Unavailability resulting from unstructured maintenance, insufficient change control & problem management, and/or untested designs or data structures among BI-reliant systems. 23 Business Layer Risks • Data store and BI is not in line with the business organization's strategic objectives and is not meeting business requirements. • Insufficient resources, training (e.g. money, people, funding) to meet business objectives • Corporate strategy not keeping pace with business needs or Business strategy not in line with Corporate Information Strategy resulting in inefficient ROI for Corporate Information Management resources. 24 Business Layer Exposures • Poor Technology ROI • Bad Business Decisions 25 Business Layer Controls • Corporate AND LOB BI Steering & Data Strategy Committees (PO1, O3) • Alignment w/Corporate Strategy (P01, O3) • Build & Model the Strategy (P01,O3) • Establish/measure milestones associated w/strategy (M1) • Track performance & errors • Training (P05, O7) • Funding (PO5, O7) 26 Administration & Operations Layer Risks • Insufficient data sources required for an effective decision support system. • Data may be accessed by persons without a business need to know. • Noncompliance with the Privacy, and Solicitation policies can lead to criticism and/or penalties from regulators and adverse publicity. Failure to have sufficient monitoring and effectiveness testing in place increases the risk that circumstances of noncompliance will go undetected. • Insufficient resources and training of team members supporting the data store can impact system availability, data integrity, BI accuracy, etc. 27 Administration & Operations Layer Risks (cont’d) • Data(BI) is not defined with regards to the owner, purpose, its use, alignment, lineage, transformation. • After a business disruption key systems are not recovered in a timely and effective manner. • Inadequate capacity due to a lack of BI utilization & performance monitoring. • Changes to BI systems and applications are made which are not authorized, tested, or appropriate. • Problems are not resolved in a timely manner. 28 Administration & Operations Layer Exposures • Poor Technology ROI • Bad Business Decisions • Inefficient database operations and analysis • System unavailability • Poor data integrity • Fraud • Unauthorized Data Access/Alteration • Disclosure of Confidential/Restricted Data • Loss of customers • Regulatory Violations Fines 29 Administration & Operations Layer Exposures (cont’d) • Processing delays • Excessive costs • Inconsistent/unreliable MIS • Untimely business decisions 30 Administration & Operations Layer Controls • Business requirements defined/documented/modeled (logical & physical models) (PO10, DS11) • Compliance w/Corporate P&P: Security, BI, IM – Encryption (DS5, PO8) – Access (DBA, User, Contractors, Archives) (DS1, 2, 5, 11, PO8) – Operations (DS11) – Change Control (AI6) – Disaster Recovery (DS4) – Archives (DS4) • Data Classification, Governance, Stewardship, & Custodianship (DS5) • Monitoring of Performance & Capacity 31 Administration & Operations Layer Controls • Training (DS5, PO8) • Metadata (DS11) – – – – – Field Definitions Lineage Transformations Models Report Logic & Fields 32 Infrastructure Layer Risks • Data transformations are not appropriate or consistent. • Incomplete, inaccurate, untimely ETL processing results in loss of data integrity. • Data Quality issues not monitored on an ongoing basis. • Lack of or poorly defined measurements/standards of what quality data includes (i.e. integrity, accuracy, etc.). 33 Infrastructure Layer Exposures • Poor data integrity • Operational Reporting Errors Ineffective/ unreliable MIS • Data Gaps • Poor Technology ROI • Bad Business Decisions • Unauthorized Data Access/Alteration • Regulatory Violations Fines 34 Infrastructure Layer Controls • • • • Transformations are modeled & tested regularly (DS11) SLA’s w/Data Stewards (DS11) Control totals/CRCs integrated into ETL (DS11) Data quality standards documented & monitored (DS11) 35 Information Services Layer Risks • Data within the ad-hoc environment is used outside of the chartered use • Data is analyzed inefficiently and/or ineffectively • Inappropriate deployment/misalignment of analytical tools with specified business need • Dashboard reports are not timely, do not have data integrity, lack consistency, do not meet business requirements • Inconsistent business definitions • Dashboard reports are not timely, do not have data integrity, lack consistency, do not meet business requirements 36 Information Services Layer Exposures • Poor data integrity • Operational Reporting Errors Ineffective/ unreliable MIS • Inappropriate data analysis conclusions • Underutilization or misuse of analysis tools\ • Bad Business Decisions • Poor ROI 37 Information Services Layer Controls • Operational report requirements documented & tested regularly (DS11) • Formal process established for requests to obtain data from the approved data store. (DS11) – Is it going to be used for production (e.g. financial reporting) • Training • Communications • Use of Analysis Tools – Access (DS5) – Purpose 38 Information Services Layer Controls (cont’d) • Monitoring & QA Testing of Analysis (PO11) • Project management P&P (PO10) • Metadata (DS11) – – – – – Lineage Purpose Definitions Structure Model • Funding – Expensive $$$$ 39 Metadata Classifications • • • • Business Process Technical Application 40 Business Metadata • • • • • • Business definitions Data structures & hierarchy Aggregation rules Ownership characteristics Subject areas Business-rule-based transformation rules • Definitions of business objectives & associated metrics 41 Business Metadata Specifics • Definitions – Data – Metrics • Models – Subject – Data • Rules – Business – Data • Data Owners & Stewards 42 Process Metadata • Origins of data Data lineage – SORs – Databases • “The When” – Schedule, frequency, & history of data captured • “The How” – Tools used for transformations – How is the data loaded 43 Process Metadata Specifics • Source/Target Maps • Rules – Transformations – Data cleansing • Audit trails – Extracts – Transformations – Loads • Data quality audits 44 Technical Metadata • Data element descriptions of: – Physical locations – Formats – Data types • Data file & table structures • Database index schemes • Implementation of data transformation rules 45 Technical Metadata Specifics • Data – Locations – Formats – Sizes – Types – Structures • Technical names • Indexing 46 Application Metadata • • • • • How data is accessed & used When data is accessed How frequent data is used Who is using the data Who is authorized to access the data 47 Application Metadata Specifics • Data access history – Who? – Frequency? – When? – How? 48 Metadata Sources • ETL & data quality tools provide: – Data transformation rules – Load statistics – Data lineage – Program dependencies – Capturing of process metadata • Data modeling tools provide – Logical & physical models – Entity and attribute definitions 49 Metadata Sources (cont’d) • Analytic tools – Application metadata • Vendor applications – Data dictionaries – Logical & physical models • Business documentation – Business policies & rules – Data ownership & stewardship – Definitions of business terms, products, processes, metrics, etc. 50 Summary • BI Background • BI Framework • BI Risks – Exposures – Controls Business Administration & Operations Infrastructure Information Services • Metadata Types & Sources Business Process Technical Application Questions?