Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

35
E-Business Sales and Purchasing Analysis using Discoverer Relational and Oracle OLAP Cameron Hawthorne Advantage BI Consulting 17 August 2010 The most comprehensive Oracle applications & technology content under one roof

description

Organisations that have deployed Oracle Discoverer reporting solutions in the E-Business Suite applications in their organisations over many years are asking how they can improve the performance of these solutions. One possibility is to add the Oracle OLAP Option and have the data pre-aggregated in cubes ready for instant analysis by the end users. This presentation will showcase a recent solution that was developed to perform Sales and Purchases data analysis using Discoverer, the OLAP Option, and Oracle Warehouse Builder.It will provide a detailed look at how to bring such a solution together and help your organisation achieve similar query performance improvements.

Transcript of Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Page 1: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

E-Business Sales and Purchasing Analysis using Discoverer Relational and Oracle OLAP

Cameron HawthorneAdvantage BI Consulting

17 August 2010

The most comprehensive Oracle applications & technology content under one roof

Page 2: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

INTRODUCTION

Page 3: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Turbo charge your reports !

Page 4: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Agenda

• What is OLAP ?• Oracle’s OLAP solutions• Case Study – Sunny Queen Farms• Need more information ?• Summary

Page 5: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

WHAT IS OLAP ?

Page 6: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

What is OLAP ?• On Line Analytical Processing

• Differs from On Line Transaction Processing

• Multi Dimensional OLAP

• Relational OLAP

• Hybrid OLAP

Page 7: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

OLAP Features

• Dimensional representation of data

Page 8: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

• Hierarchy definitions for dimensions

Sales Channel Hierarchy

OLAP Features

Channel Dimension

All Channels Level

Class Level

Channel Level

Page 9: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

OLAP Features

• Pre-Summarised data

Customer – Ship To CodeProduct – SKU Code

Time - Day

Page 10: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

OLAP Advantages• Faster reporting of information – query response time

• Business like representation of data to users

• Fast display of rotating dimensions

• Drill down from summarised totals through hierarchies

• Analytical features

Page 11: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

OLAP Trade-offs• Additional resources to pre-summarise data

– Time, Disk, Memory

• Management of data sparsity– Cube design, composites of dimension values

• Additional movement of data– From data warehouse or mart to cubes

Page 12: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

ORACLE’S OLAP SOLUTIONS

Page 13: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Oracle’s OLAP Solutions• Relational OLAP

– Fact Tables, Dimensions, Materialized Views– Analytical SQL (Rollup, Cube, Group By, Rank etc)

• Multi-Dimensional OLAP– Oracle Express Server– Oracle Database Enterprise Edition OLAP Option– Oracle Hyperion Essbase

• Hybrid OLAP– Cube Organized Materialized Views

Page 14: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

OLAP Option Features• Database Enterprise Edition Option

• The OLAP engine runs within the Oracle Database

• Dimensional objects are stored in their native multidimensional format

• Data security is administered in the standard way

• Applications can query dimensional objects using SQL

Page 15: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

OLAP Option Features• In the database

– OLAP DML• Functions, Commands, Programs

– Managing AW’s with DBMS_AW Package– Security with grant and revoke– Database backups, flash recovery

• Oracle Warehouse Builder Integration• Analytical Workspace Manager

Page 16: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

OLAP Option Features• Data Exploitation Software

– Oracle Business Intelligence 11g– Oracle Discoverer Plus for OLAP– Oracle OLAP Excel Add-In, Simba MDX– Oracle BI Beans – J2EE development– Oracle Application Express– Any SQL Reporting Tool, using 11G OLAP Views

Page 17: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

CASE STUDY – SUNNY QUEEN

Page 18: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Sunny Queen Background

• Sunny Queen Farms– 100% Australian Owned company supplying high

quality farm fresh eggs to consumers in all states– Implemented Oracle E-Business Suite in 2001– Selected Oracle Sales Analyzer as reporting solution

• Oracle Express Server database• Windows and web clients• Sales, Rebate, and Case Deal data

Page 19: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

SQ Reporting History• 2002 – Changed hierarchies, added budget sales data• 2003 – Added Purchasing actual data analysis• 2004 – Acquired new business – Egg and I, Started archiving• 2005 – Added Purchasing forecast data, changed rebate data focus

to promotional analysis• 2006 – Upgraded to latest version of OSA, changed data model –

new hierarchies and definitions• 2007 – Replacement Discoverer for OLAP solution using Oracle

OLAP Option 10gR1

Page 20: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

SQ 2008 Problems• Business model and staff changes over the years led to

data model change requests• Discoverer for OLAP issues:

– Missing OSA functionality– OLAP 10gR1 software bugs causing significant outages– Still relying on OSA during outages– OSA out of support

• Everything on the 1 box– Out of date OS and software patches

Page 21: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

2008 OSA Architecture

Page 22: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

2008 Disco OLAP Architecture

Page 23: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Current Architecture

Page 24: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Additional details• Data drawn directly from E-Bus schemas

– Requires a knowledge of schemas and tables– Not an E-Bus upgrade-proof solution

• Work-arounds required to meet OWB Basic Edition rules– No Process Flow

• Shell scripting that calls OWB APIs to execute tasks

– No OLAP design integration• Use Analytical Workspace Manager to design and build cubes

Page 25: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Sales Analysis• Sales measured in $$$ and Volume (Units, UOM, Kgs)• Dimensions

– Product – 5 hierarchies incl. Brand, Segment, Size– Geography – Territory and Customer Ownership structures– Channel – Retail, Route Trade, Foodservice– Warehouse – SQ Distribution Centres– Time – Financial and Calendar Years

Page 26: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Purchasing Analysis• Purchases in $$$ and Volume (Units, UOM, Kgs)• Dimensions

– Product – 5 hierarchies incl. Brand, Segment, Size– Grower - Ownership structure– Warehouse – SQ Distribution Centres– Time – Financial and Calendar Years

Page 27: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Margin Analysis• Gross Margin

– Sales $$$ - Purchasing $$$

• Net Margin– Gross Margin $$$ - Estimates of other variables costs $$$

• Needs common dimensions– Product, Warehouse, Time

• Assumes product is sold in same timeframe it is sourced

Page 28: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Variance Analysis• Sales Budgets vs Actuals

– Budgets loaded from .CSV files into Application Express app– Financial estimates of sales performance at high level

• Purchasing Forecasts vs Actuals– Forecasts keyed into Application Express app– Grower estimates at high level

• Provides managers with performance in context

Page 29: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Benefits and Costs• Benefits

– Improved query response times– Stable environment, minimal outages– Minimised impacts on main transaction system– Data model that met current business requirements

• Costs– Additional software licensing– Additional hardware costs

Page 30: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

More information

• Oracle OLAP on Oracle Technology Network– www.oracle.com/technetwork/database/options/olap

• Oracle by Example– www.oracle.com/technology/obe/11gr1_db

• Mark Rittman’s blog– www.rittmanmead.com/blog/

• Oracle User Group BI SIG– www.oraclebiwa.org

Page 31: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

SUMMARY

Page 32: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Summary

• What is OLAP ?– OLAP Features

• Oracle’s OLAP software– Enterprise Edition OLAP Option

• Sunny Queen Farms Case Study

Page 33: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco
Page 34: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Contact InformationFor more information about today’s presentation, please contact:

Cameron Hawthorne

Managing Consultant

Advantage BI Consulting

W: www.advantagebi.com.au

E: [email protected]

Page 35: Cameron Hawthorne - E Bus Sales analysis with Olap and Disco

Tell us what you think…

• http://feedback.insync10.com.au