1 Business Intelligence: Effective Decision Making Bellevue College Linda Rumans IT Instructor, Business Division Bellevue College
2 Current Status What do I do??? How do I increase sales???? How do I make my product better??? Business Users Mountains of Data
3 Mountains of Data From Operational Systems ERP (Enterprise Resource Planning) Sales/Order Inventory Customer Relationship Management (CRM) Web Sites Orders Click-stream
4 Mountains of Data Organizations have lots of data Data is not in a form that is useful to decision-makers Not easy to review Not informative nor insightful
5 Today s Information Flow Business in 90 s invested in transactional systems: Supply Chain Management (SCM) Customer Relationship Management (CRM) Enterprise Resource Planning (ERP) Manufacturing Resource Planning (MRP) Finance (budget, forecasting and reporting)
6 Proliferation of Data Sales Procurement Operations Finance Reporting Layer Transaction Layer CRM SCM MRP Finance Silos of data by functional area
7 Data from Disparate Sources Sales Sales Sales Sales Reporting Layer Transaction Layer Div 2 Region: B Region: A Div 1 Silos of data within large organizations
8 Business Intelligence Business is now investing in Business Intelligence Business Intelligence is about making effective business decisions
9 What is BI? The process by which an organization manages large amounts of data, extracting pertinent information, and turning that information into knowledge upon which actions can be taken.
10 What is BI? Business intelligence (BI) is a broad category of application programs and technologies for gathering, storing, analyzing, and providing access to data to help enterprise users make better business decisions.
11 BI Involves PEOPLE and Technology Involves using a rational approach to management Involves a continuous cycle of measurement, adjustment & re-measurement
12 The BI Cycle BI start
13 Reasons for BI BI enables organizations to make well informed business decisions and gain competitive advantage. BI enables organizations to use information to quickly and constantly respond to changes.
14 Benefits of BI Improved performance based upon timely and accurate information Elimination of guesswork Expedited decision making Early visibility of changes: Customer buying patterns Supply chain activity Financial arrangements
15 Benefits of BI Single Version of the truth Accurate, timely data available to all levels of the organization
16 To Note: Although we call it Business Intelligence, the concepts and techniques are applicable to almost any organization including those in health care, biotech, education, government
17 BI Activities BI applications include the activities of: decision support, query and reporting, online analytical processing (OLAP), statistical analysis, forecasting, and data mining.
18 BI Users There are many different users who can benefit from business intelligence Executives Business Decision Makers Information Workers Line Workers Analysts
19 BI Solutions- How to make it happen Two main components: Data Consolidation and Storage Data Retrieval, Analysis and Presentation
20 BI Curriculum Multi-Dimensional Analysis Data Warehousing Data Mining Dimensional Modeling Data Visualization
21 The Problem How do I retain customers? How do I increase sales???? GAP How do I make my product better??? Business People Mountains of Data
22 Bridging the Gap Need data storage structures to facilitate fast analysis of huge volumes of data Need software to provide access to the data, allow flexible manipulation, and provide meaningful presentation
23 Data Storage Structures Multi-Dimensional Databases Cubes
24 Multi-Dimensional Databases Measures Any quantitative expression Some are designated as Key Performance Indicators (KPI) Appropriate to the business process. Dimensions How we describe the measures: Product/Customer/Region/Time These are the By s What were our Customer Sales by Product Line by Region by Quarter for the past two years?.
25 Logical Structure
26 Multi-Dimensional Databases (Cubes) Multi-Dimensional Database (Cube) Business Intelligence Programs Data Warehouse ODS ODS ODS * ODS = Operational Data Store Relational Database Programs
27 Multi-Dimensional Databases Multi-Dimensional Database (Cube)
28 Software Applications Business Person Business Person Reporting Applications Analytic Applications Multi-Dimensional Database (Cube) Business Person Score Cards Dashboards
29 Analytics Reporting Applications Limited user interaction Fulfill a significant portion of an organization s information needs Analytic Applications Allow users to visualize and explore data following their train of thought Extensive interactivity
32 Analytic Application
33 Summary Students learn to: Create multi-dimensional databases Create professional quality reports Use analytics to provide in-depth data analysis
34 Data Warehousing Designing a Data Warehouse
35 Data Warehouse Topics Decision Support Systems history Requirements Gathering Where data located, owners, definition, how often updated Data Analysis Determine for table structures
36 Data Warehouse ETL Processes & Deliverables Cleaning & Conforming Valid, missing Address, gender Schemas Dimension Tables Fact Tables
37 Data Consolidation & Storage Customers Sales Procurement Suppliers Operations Finance Shared Reporting Shared Data Layer Data Warehouse Transaction Layer SCM CRM MRP Finance Operations and financial information is shared across the organization from same core data
38 Data Warehouses Multi-Dimensional Database (Cube) Data Warehouse ODS* ODS ODS *ODS = Operational Data Store
39 How is data consolidated? This is difficult!!!!! Data is often spread across multiple systems, stored in different formats, and may even be localized for different countries
40 Transforming Data Data must be transformed for consistency and meaning Transformations may be as simple as copying columns or may be incredibly complex Common transformations include: Hard-coded changes ( T to 1) Looking up values in a table (mapping a customer number across disparate systems) Inserting dummy records and mapping them to unknowns (inserting an Unknown customer)
41 Cleansing Data Data must be cleansed to be meaningful All companies have bad data in their systems Data may be missing Data may be inconsistent Data may be wrong
42 Data Warehouses ETL (extract, transform and load) processes are needed to create data warehouses This is an arduous and technical process that can account for a large percentage of a BI project cost!!!!
43 Data Mining
44 Data Mining The process of identifying patterns in data Goes beyond simple querying of the database Goes beyond multi-dimensional database queries as well
45 Data Mining Data Mining works for problems like: Develop a general profile for credit card customers Differentiate individuals who are poor credit risks Determine what characteristics differentiate male & female investors.
46 Data Mining vs. Data Query Use data query if you already almost know what you are looking for. Use data mining to find regularities in data that are not obvious.
47 Data Mining Applications Fraud detection Targeted Marketing Risk Management Business Analysis
48 Origins of Data Mining Mathematics Statistics Numerical Analysis Artificial Intelligence/Machine Learning Computer Science Data Storage and Manipulation
49 How does Data Mining work? Uses induction-based learning: The process of forming general concept definitions by observing specific examples of concepts to be learned.
50 How does Data Mining work? What-Cha-Ma-Call-Its NOT What-Cha-Ma-Call-Its
51 How does Data Mining work? Which of these are What-Cha-Ma-Call-Its?
52 Data Mining Process List of Customers: -some bicycle buyers -some not Data Mining Software Model List of Prospective Buyers Model List of Likely Buyers
53 Overview of Mining Strategies Note: This representation is over-simplified and data mining strategies are continually being invented.
54 More on our Curriculum
55 Skills Written communication Problem Solving Analytical Troubleshooting Software Microsoft SQL Server Management Studio SQL Server BI Development Studio SQL Server Reporting Services Pro Clarity
56 Delivery Methods Online: Distance Education, reaches wider market Telecourse: tremendous effort to create, but once created easy to deliver Televised, DVDs, online for homework, exams Hybrid: Meet once a week, the rest online On campus: evenings only
57 Delivery Methods Use of Camtasia for Software demonstrations PowerPoint lectures Pod casting
58 Certificates Business Intelligence Analyst (5 classes) Multi-dimensional analysis, data warehousing, data mining, statistics, general business 2 quarters full-time/ 3 quarters part-time Business Intelligence Developer (4 additional classes) Dimensional modeling, data visualization, multidimensional II, data warehousing II (more programming with SQL Server) Web site:
59 Certificates Relational Database Analyst (6 classes) SA & D, programming, reporting, spreadsheets, db theory 2 quarters full-time/ 3 quarters part-time Relational Database Developer (3 additional classes) Programming, SQL, group processes Web site:
60 Jobs Business Analyst Data Analyst Functional Analyst Marketing Analyst
61 Jobs Report Developer Data Modeler ETL Developer Data Architect Data Warehouse Designer Data Warehouse Developer Data Warehouse Administrator Database Administrator
62 Jobs Business Intelligence Consultant Business Intelligence Developer Business Intelligence Analyst Business Intelligence Project Team Member
63 Jobs One of the fastest growing segments of IT Less likely to be outsourced May exist in business units rather than IT Knowledge/understanding of the organization is key