Transcript Extending Financial Reporting out of Oracle

Extending Financial Reporting out of
Oracle Financials Using Hyperion
Essbase
George Cooper
Hackett Technology Solutions
Hackett Technology Solutions
• Founded in April 1997
– Traded on NASDAQ as ANSR
• World’s leading repository of
enterprise best practices
strategies and metrics
– Benchmarking and Advisory Services
– Business Transformation (HR, IT, Planning and
Performance Measurement, Finance, Supply Chain )
• Over 700 professionals with 12+
years experience across more
than 25 competencies
• Consulting and system
integration expertise with
comprehensive skills in:
– Oracle, SAP
– Hyperion
Hackett’s Oracle Practice
YEARS OF
EXPERIENCE
– Certified Oracle Implementation
Partner since 1995
– Completed over 350 Oracle and
PeopleSoft implementations,
upgrades and optimization projects
– Implementing Oracle Solutions based — Highly skilled consultants
on Hackett-Certified™ Practices
— Large-scale Oracle
15 +
implementation experience
35%
— Average 7 years Oracle
10
15
35%
experience, 12 years business
experience
5 - 10
25%
— 78% are Hackett-Certified
5% Average 12 years experience
Advisors
— Many of Oracle’s Leading
Accounts Have Already
Discovered the Benefits of the
Answerthink / Hackett Group
Approach
Hackett’s Hyperion Analytics Practice
• Strategic and technical implementation consulting
• #1 Hyperion Americas Reseller Award at Solutions 2006
and 2007
• Member of Hyperion Partner Advisory Council; Customer
Advisory Boards and participant in the Partner
Development Exchange
• 400+ successful Hyperion projects, with well over 100 in
2006 and 50+ of those System 9
• 100+ resources dedicated exclusively to Hyperion
• Scores of Hyperion-certified consultants, plus Preferred
Partner Certification
Hyperion Clients
Agenda
•
•
•
•
Hyperion Essbase Overview
Essbase Technical Overview
Typical Implementation
Case Studies
Hyperion Essbase Overview
What is Essbase?
• The leading enterprise multi-dimensional database
engine
• Provides real-time analytic infrastructure for business
intelligence and enterprise performance management
(EPM) applications.
• Engineered for scalability, security, and rapid-response.
• Through an intuitive interface, business users can
manipulate large data sets to model complex scenarios,
forecast outcomes, and perform “what-if” analyses to
identify trends and optimize business results.
Oracle Hyperion Essbase Datasheet
What is Multi-Dimensional?
• Uses a cube metaphor to
describe data storage.
• An Essbase database is
considered a “cube”, with each
cube axis representing a
different dimension, or slice of
the data (accounts, time,
products, etc.)
• All possible data intersections
are available to the user at a
click of the mouse.
Multi-Dimensional vs. Relational
• Multi-dimensional database are
usually queried top-down – the
user starts at the top and drills
into dimensions of interest.
Total Products
P01
P02
P03
P01
P02
P03
– Can perform poorly for transactional
queries
• Relational databases are usually
queried bottom-up – the user
selects the desired low level data
and aggregates.
– Harder to visualize data; can
perform poorly for high-level queries
Total Products
Why Use Essbase?
• Rich User Experience – users “converse” with
the data
• Business and Finance can manage their own
metadata and reports
• Highly advanced calculation engine
• Easy integration of data sources, including
manual input
• Large scalability
• Robust, cell-level security
• Many sophisticated reporting tools
Rich User Experience
• Sub-second response
• Intuitive interface, especially with Microsoft
Excel
• Powerful adhoc analysis that allows users
to query virtually any database intersection
in seconds
• Visually understand the relationships in
the data
• Easily built reports without IT involvement.
Rich User Experience
Excel
Essbase
Demo
Business Metadata Management
• Graphical administration console allows
authorized administrators to “see” their
data structures
• Simple specification of alternate rollups for
specialized reporting
• Allows the application to evolve as quickly
as the business
Business Metadata Management
Essbase
Administrative
Services (EAS)
Demo
Powerful Calculation Engine
• Over 350 built-in functions, including:
– Financial functions, such as net present value, rate of
return, and compound growth
– Custom multi-dimensional functions such as
@ALLOCATE to drive data to multiple business
intersections
– Complete time-series support
– Support for summary-level input
• Both run-time and batch calculations
• MDX support
Integration of Data Sources
• Unique multi-user read/write technology
• Information from many data sources can
be easily integrated into one database,
and thus one set of user queries
• In particular, allows budgets and forecasts
to be fully integrated with actuals
Technical Overview
Data Storage
• Multi-dimensional, with own proprietary
storage
• Two primary data storages:
– Block Storage Option (BSO) – Record-based
storage; supports write-back and features the
batch calculation engine.
– Aggregate Storage Option (ASO) – Cellbased storage; supports fast aggregation with
a large number of dimensions.
Block Storage Aggregations
In general, all members combinations are calculated
during an aggregation
– Can be optimized for faster performance
– Allows complete control of calculations
ASO Storage Aggregations
Engine decides which level intersections should be
calculated to minimized retrieval time
– Specific aggregations can be specified to optimize particular
queries
Smaller Agg. Size
Larger Agg. Size
Scalability
• BSO databases generally are impractical with more than
five or six hierarchical dimensions, depending on the
number of members, depth of hierarchies and structure
of the data
• ASO databases have no set limit on number of
dimensions – 20+ dimensions are possible
• Both storage types support:
– Hundreds of thousands of outline members
– Attribute dimensions, which are based on a one-to-many
relationship with the base members of another dimension
• No additional storage or calculation time is needed for
Attribute dimensions in BSO
– Drill-through to relational detail
Partitions
• Partitions are dimension slices that are shared
between Essbase database
– Replicated – The data is physically transferred
between the source and target cubes
– Transparent – The data in the source cube is queried
at retrieval time
• Process is seamless to the user – all data
appears to be in the target cube
• ASO and BSO databases can be linked together
with a transparent partitions
Typical Implementation
Partnership with Oracle GL
• Users typically use Essbase for:
–
–
–
–
–
–
Historical and forecasted performance
Budget variances
Variance and profitability analysis
Performance trends
Profitability metrics and Foreign Exchange impact
Generally any query using non-transactional data
• Users typically use Oracle Reports for:
–
–
–
–
Viewing individual transactions
Auditing GL entries
External reporting
Generally any query that needs transaction data
Typical Support Model
• IT supports core functionality:
– Servers and software installation
– Daily extracts and loads from Oracle Financials and other data
sources
– Core calculation scripts and database dimensionality
– Essbase automation
• Central Administration supports:
– Security
– Primary hierarchies and metrics
• Line Finance or Business supports:
– Most report development
– Alternate hierarchies for specialized reports
– Budget and forecast input and review
Basic Implementation
Hyp Planning
Forms
Excel
Templates
Oracle
Financials
Hierarchy
Metadata
Other Data
Sources
Actual Balances by
Month / GL Segment
Text Files
Manual
Forecast & Budget
Input
Essbas
e Admin
Service
s
Essbase
Finalized Budgets
Adhoc
Excel
Queries
Excel
Reports
Dashboards
& Production
Reports
Oracle
Upload
Implementation with Master Data Mgmt
Hyp Planning
Forms
Oracle
Financials
Other Data
Sources
Actual Balances by
Month / GL Segment
Excel
Templates
Text Files
Manual
Forecast & Budget Input
Master Data
Mgmt
(MDM)
Essbas
e Admin
Service
s
Essbase
Finalized Budgets
Adhoc
Excel
Queries
Excel
Reports
Dashboards
& Production
Reports
Oracle
Upload
Case Study #1
Budgeting and Forecasting for a
Large Retail Company
Problem
Build a budgeting and forecasting model that
can support the needs of the individual brands
while minimizing both IT support and the
technical expertise required by the Finance
administrators
Challenges
• Separate applications could provide the flexibility
required by the individual brands but would increase
development time and ongoing technical support
• Reporting requirements are significantly different in each
brand
• A single application would reduce technical support but
negatively impact planning flexibility and possibly reduce
performance and reliability
• The primary business hierarchies must stay in sync with
all brands
• Planning administration in each brand must require a
minimum of training since turn-over is relatively high in
the finance groups
• A combined corporate view of actuals, budgets and
forecasts must be supported
Solution
• Have separate physical Essbase databases for each
brand while having a commonly maintained outline,
calculation scripts, automation procedures and Excel
utilities
– Keeps each brand’s data physically separate to reduce
performance risk while increasing security
– Central outline and core functionality reduces maintenance,
increases reliability and minimizes training for each brand
administrator
– Central maintenance of core hierarchies and metrics keeps one
version of the truth
• Include functionality in the database to allow any account
to either have direct input or be calculated as a
percentage of a selected driver (revenue, headcount,
etc.)
– Allows the brands to decide how each account is calculated
Solution (cont)
• Task each brand administrator with building their own
reports, input templates and alternate hierarchies
– Gives each brand the flexibility to internally report their business
in the manner their management desires to see it
– Brands can share reports and templates where appropriate
• Build a central administration console to automate
common administrative functions
– Calculating input data, maintaining scenarios and controlling the
budget process in a single, easy to use interface
• Build a separate corporate database with a replicated
partition to the individual brand databases
– Supports a combined view of the corporation
Architecture
Hierarchy
Metadata
Oracle
Financials
Actual Balances by
Month / GL Segment
Common Automation
Outline and Data Updates
Brand A
Essbas
e
Excel
Templates
Brand B
Essbas
e
Manual
Forecast & Budget Input
Brand C
Essbas
e
Essbas
e Admin
Service
s
Central
Essbas
e
Outline
Replicated
Partition
Corp
Essbas
e
Case Study #2
Budgeting and Forecasting for a
Large Trade Show Management
Company
Problem
Build a budgeting and forecasting model that
can support very detailed forecasts and
budgets while providing quick and transparent
access to all data
Challenges
– Large number of business dimensions
potentially increase database size and
calculation time
– Planners are spread over a wide geographic
area with little technical expertise
– A very large number of individual projects
need to be forecasted
Solution
– Use Hyperion Planning as the overall planning engine
• Web-based planning input
• Finance administrator can easily define forms
• Little training needed for budget / forecast input
– Keep current data in the Planning BSO database but
move actuals and historical plans into a separate
ASO database
• BSO database allows write-back for maintaining the current
information
• ASO database allows fast loading and aggregation times for
the much larger volume of historical and actual data
– Optional: Link the two database via a transparent
partition to give the users one view of the data
Architecture
Oracle
Financials
Actual Balances by
Month / GL Segment
Hierarchy
Metadata
Hyperion
Planning
Current
Essbas
e
Hyperion
Planning
Admin
Console
Hyperion
Planning
Web Forms
Level0 Export
ASO
History
Essbas
e
Excel
Reports
Architecture (Alternative #1)
Oracle
Financials
Actual Balances by
Month / GL Segment
Hierarchy
Metadata
Hyperion
Planning
BSO
Input
Essbas
e
Hyperion
Planning
Admin
Console
Hyperion
Planning
Web Forms
Level0 Export
ASO
Report
Essbas
e
Excel
Reports
Architecture (Alternative #2)
Hierarchy
Metadata
Oracle
Financials
Actual Balances by
Month / GL Segment
ASO
Actual /
History
Essbase
Hyperion
Planning
Planning
Essbase
Hyperion
Planning
Admin
Console
Hyperion
Planning
Web Forms
Transparent
Partition
Excel
Reports
Case Study #3
ERP Reporting at a Large
Semiconductor Company
Problem
Build a suite of reports from various ERP
modules during a new Oracle implementation
under a very aggressive time frame
Challenges
– Development time frame was short and
business involvement constrained during the
ERP implementation
– Users required the flexibility to design and
build reports and dashboards over time but
see the data immediately at go-live
– Some modules (such as quality assurance)
had a large volume of data and a large
number of dimensions
– Data had to be refreshed nightly
Solution
– Build separate Essbase databases for each ERP
module with the dimensionality and level of detail
necessary to manage the business
– Task the business and finance staff to develop their
reporting through Excel SmartView
– For each application combine a BSO Essbase
database for core calculations and an ASO Essbase
database for user reporting
• Supports a large number of dimensions and extremely large
data volumes
– Develop formatted reports in Hyperion Financial
Reports and Hyperion Web Analysis where
appropriate
– Use Hyperion System9 Interactive Reporting for
transaction-level reports
Architecture
Oracle
ERP
Hierarchy
Metadata
Excel
Reports
Text Extracts
BSO
Level 0
Essbas
e
Calculations
Level0
Export
ASO
Report
Essbas
e
Dashboards
& Production
Reports
Contact Information
Michael Cook
Director of Business Development
Phone – 415.435.0344
[email protected]
George Cooper
Manager, BI Northern California Practice
Phone – 510.290.9538
[email protected]