Description
Choosing a Business Intelligence (BI) offering is an important decision for an enterprise, one that will have a significant impact throughout the enterprise.
Evaluating Business Intelligence Offerings:
Business Objects and Microsoft
www.symcorp.com
August 2, 2005
Evaluating Business Intelligence Offerings:
Business Objects and Microsoft
Table of Contents
Introduction.................................................................................................................................... 3
What is Business Intelligence...................................................................................................... 3
Key Considerations in Deciding on a BI Offering ...................................................................... 4
Requirements of Business Users................................................................................................. 4
Requirements of Application Developers..................................................................................... 4
Requirements of IT Professionals................................................................................................ 5
Summary of Key Considerations ................................................................................................. 5
Comparing the BI Strategies of Business Objects and Microsoft............................................ 5
A Historical Overview of the BI Offerings of Business Objects.................................................... 6
A Historical Overview of the BI Offerings of Microsoft ................................................................. 7
An Evaluation of the BI Offering of Business Objects .............................................................. 8
An Evaluation of the BI Offering of Microsoft ............................................................................ 9
Summary ...................................................................................................................................... 10
About Symmetry Corporation .................................................................................................... 10
2 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings:
Business Objects and Microsoft
Introduction
Choosing a Business Intelligence (“BI”) offering is an important decision for an enterprise, one
that will have a significant impact throughout the enterprise. The choice of a BI offering will affect
people up and down the chain of command (senior management, analysts, and line managers)
and across functional areas (sales, finance, and operations). It will affect business users,
application developers, and IT professionals. Before deciding on a BI offering, business decision
makers must evaluate a company’s BI offering with respect to what it has to offer each of these
stakeholders. This paper will analyze the key considerations for deciding on a BI offering, review
the BI strategies of Microsoft and Business Objects, and then evaluate each company’s offerings
in light of these key considerations.
What is Business Intelligence
BI is a broad topic, covering many different functions (e.g., reporting and analysis) and
technologies (e.g., data warehouse, OLAP, portal). An examination of the literature shows many
varying definitions of BI. These definitions fall into two classes: a technical description of the
components that make up a BI solution and an explanation of the business purpose of BI. A good
definition of BI that encompasses both technical functionality and business purpose is the
following: BI is a broad category of applications and technologies for gathering, storing, analyzing,
and providing access to data to help enterprise users make better business decisions. To
effectively evaluate the offerings of BI vendors, how well an offering satisfies both sides of the
equation must be examined.
BI applications include the activities of decision support systems (DSS), query and reporting,
online analytical processing (OLAP), statistical analysis, forecasting, and data mining. Another
way of phrasing this is that BI applications take data that is generated by the operations of an
enterprise and translate that data into relevant and useful information for consumption by people
throughout the enterprise.
To deploy BI successfully throughout an enterprise requires a platform, not a hodgepodge of tools
and technologies. As such a BI platform must be able to address all the BI needs of an
enterprise. Gone are the days where BI was a point solution used by a limited set of highly
trained analysts. In today’s market, BI is pervasive throughout an enterprise and in many cases
mission critical. As such, a BI platform must be able to address both the breadth and depth of the
needs of an enterprise.
3 of 10
3 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
Key Considerations in Deciding on a BI Offering
Business decision makers must evaluate both the breadth and depth of a BI offering in order to
make an appropriate decision on a BI offering. The breadth of a BI offering is a measure of how
well the BI offering supports the different requirements of the BI stakeholders. The depth of a BI
offering is a measure of its vertical integration, i.e., how well the BI offering enables a business to
take raw data from a production system and transform it into relevant and useful information and
then deliver this information to business users in the proper context.
There are many steps required to generate consumable BI from raw data and there are various
BI stakeholders within an enterprise that have a vested interest in the process. IT professionals,
application developers, and business users (who are also the consumers of BI) all play a role in
the development of a BI solution. Business users define the business rules that determine how
the raw data must be transformed. Application developers develop the processes for acquiring,
consolidating, and presenting the raw data based on the business rules. IT professionals manage
the processes, ensure availability, and enforce security.
Requirements of Business Users
Different types of business users have different requirements of a BI offering.
• Analysts – Analysts support managers with performance management analysis. Analysts
require a powerful and interactive environment that allows them to create metrics and
navigate the data in an ad-hoc setting. This type of user requires tools for analytics, statistics,
predictive modeling, and advanced visualization.
• Managers – Managers at all levels need BI to assist them in making informed business
decisions. This type of business user requires a friendly query environment that also supports
the ability to generate ad hoc reports and delivery mechanisms that enable managers to
disseminate information up and down the chain of command.
• Operations workers – Operations workers use BI as part of solving a larger issue. For
example, as part of servicing a customer, a retail clerk might recommend other related
products to a customer. This type of worker requires BI that is embedded in a production
application, rather consuming BI as part of a BI application.
Business users of all types want to reduce their dependence on IT, but still have confidence in the
numbers, have advanced analytics, superior query performance, and access to timely information
in the format and delivery mechanism of their choice, whether through a portal, a spreadsheet, or
email. Satisfaction of these business user requirements enables BI to truly become mission
critical, fulfilling the promise of BI, and providing businesses with competitive advantage in the
global marketplace.
Requirements of Application Developers
Application developers must be able to develop the variety of BI application types required by
business users that are essential to enable enterprises to obtain competitive advantage in the
global marketplace. The range of capabilities that a BI offering needs to support is as varied as
the BI applications required by enterprise business users. BI applications such as sales analysis
need to be able to handle large data sets (terabyte) with very long lists of dimension members (in
the millions). Other BI applications must support complex calculations for the derivation of key
performance indicators or financial reporting modeling. Other BI applications merge BI analytics
with data collection for budgeting, planning, and forecasting. Still other BI applications require
very low data latencies for use in business activity monitoring applications to create real time BI.
The BI applications developer requires a BI offering that is capable of supporting this entire range
of BI applications.
In an enterprise today, BI is frequently embedded in business processes that support operations
workers and needs to be seamlessly integrated into existing applications, and then easily
extended as new BI needs are discovered. Application developers must be able to use existing
4 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
skill sets, reuse code and components, and leverage existing applications and infrastructure to be
able to meet the increasing need for BI that can be easily maintained and extended without
requiring new skill sets. A crucial factor for the productivity of application developers is having a
single development environment that allows them to work on all aspects of BI, from the data to
the analytics to the user interface using a single development language, and that supports team
development.
Further, application developers must be able to easily extend BI applications using third party
tools and technologies where necessary. These third party products must integrate existing BI
applications, rather than operate parallel to the platform. This requirement is an acceptance of the
fact that there is no perfect product in the market. A BI offering might fulfill most of an enterprise’s
needs, but there are always a few cases where requirements cannot be met with the existing
offering. Extensibility offers a safety valve in such cases.
Requirements of IT Professionals
IT professionals require a BI offering that enables them to deliver mission critical BI; namely a BI
solution that is highly available, reliable, and secure. IT professionals require a back-end solution
that is fault-tolerant and scalable; that supports change control and scriptable deployment; and
that enables them to leverage their existing resources and skill sets while building on the current
IT platform and infrastructure. Furthermore, IT professionals must be able to deliver real time or
near real time data to business users with minimal degradation in query performance.
Summary of Key Considerations
The scope of BI uses is large enough that trying to meet all of the needs discussed above with a
portfolio of stand-alone tools is inefficient, costly, and ultimately ineffective. Yet in order to provide
leverage, the BI offering must have a consistent manner of accomplishing tasks while integrating
with the existing infrastructure. This consistency will result in reduced training cost for business
users, application developers, and IT professionals; shorten development time for new
applications and enhancements; and result in greater acceptance by business users.
In summary a BI offering must:
? Support the end-to-end building of a BI application, from data source to the presentation
of information
? Encompass the diverse business user needs of analysts, managers and operations
workers
? Provide a solution for the varied application areas such as finance, sales, and production.
? Address the needs of various BI stakeholders in the enterprise: business users,
application developers, and IT professionals.
Comparing the BI Strategies of Business Objects and Microsoft
Business Objects and Microsoft have very different strategies for approaching the BI market,
which has resulted in very different BI offerings.
Business Objects started with a client tool for business users in 1990 and then pursued a
backwards integration strategy to provide a more complete BI platform for the IT professional and
application developer arena with its BI Application Foundation for analytical application
development. The strategy of Business Objects has been a strategy of acquisitions to evolve into
a broader back-end platform while still maintaining its focus on its original end-user audience.
Microsoft entered the BI market in 1997 with a high-performance analytical server offering and
pursued a forward integration strategy to provide a more complete BI platform based on an open
architecture supporting both Microsoft and third-party developer and end-user tools compatible
5 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
with its analytical server offering. Because Microsoft chose to enter the BI market 7 years after
Business Objects, the initial technology base for the Microsoft product reflects a more modern
architecture. The Microsoft BI architecture also reflects Microsoft’s strategic focus on a wide
audience of IT professionals, application developers, and business users.
A Historical Overview of the BI Offerings of Business Objects
Business Objects’ entry into the BI marketplace in 1990 was a client tool that displayed data from
relational databases in tables and graphs before such functionality was considered BI. Business
Objects differentiated itself in the marketplace by providing a semantic layer (called a “Universe”)
that insulated the business user from the physical database structure details. A Universe is
basically a way to rename columns to more appropriate business terms, create new calculated
columns, and hide the complexity of table joins.
In 1997, to meet the growing need of web based deployments, Business Objects introduced Web
Intelligence, a browser-based query tool. While the initial functionality of this tool was limited,
Web Intelligence is now the premier query tool within the Business Objects product line.
As the concept of BI took hold in the market, Universes were enhanced to represent dimensions
with hierarchies and measures. Also, the client side was enhanced to provide better BI-style
analysis, such as top x queries.
As Business Objects recognized the growth in OLAP multidimensional databases, Business
Objects purchased OLAP@Work to obtain both an Excel-based query tool and a vehicle for
access to Microsoft’s Analysis Services. OLAP@Work became the basis for BusinessQuery.
In 2001, Business Objects released Application Foundation, an analytical application framework
for jumpstarting analytical applications. Its Dashboard was initially part of the Application
Foundation and has since been split out as a separate offering. The Application Foundation
contains three analytical engines:
• Set analysis for segmentation of various business perspectives such as customer or
suppliers
• Predictive analysis for time series forecasting
• Statistical process control for Six Sigma and Total Quality Management.
Despite its name, the Application Foundation provides features more oriented to end-users than
IT professionals and application developers, which is a reflection of Business Objects’ traditional
orientation toward the end-user.
Next, Business Objects acquired Acta, a data integration vendor for Data Integrator, its extraction,
transformation, and loading (ETL) product. Data Integrator has failed to gain market traction and
has languished since its acquisition by Business Objects. In 2003, Business Objects acquired
Crystal Decisions, which gave Business Objects a managed reporting solution with Crystal
Enterprise and Crystal Reports, and an OLAP query and analysis tool in Crystal Analysis (which
is now known as OLAP Intelligence).
Business Objects has filled out its product line mostly through an acquisition strategy. While this
strategy has allowed Business Objects to rapidly expand its product line, this rapid expansion
strategy has come at a price. Business Objects has experienced difficulties in integrating
companies located in diverse geographical areas with different business structures and cultures,
and has not always been successful in this integration, as in the case of Acta. The acquisition of
diverse products has also provided distinct integration challenges. The most recent release
Business Object XI has repackaged the entire Business Objects suite of products. Before
Business Object XI, the Crystal and Business Objects product lines were separate and distinct.
With this recent release, Business Objects has brought common administration to the product
6 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
line, though the tools themselves remain discrete. Furthermore, Web Intelligence and Crystal
Reports still have different file formats.
A Historical Overview of the BI Offerings of Microsoft
Microsoft entered the BI marketplace in 1997 with the release of SQL Server 7. There were two
aspects to Microsoft’s initial foray into BI. The first aspect was OLAP Services, which was based
on technology that Microsoft acquired from Panorama Software. The second aspect was Data
Transformation Services (DTS), a data integration tool. OLAP Services was well-received
because it lowered the cost of entry for analytics and raised the bar for performance and
scalability in a multidimensional database. Microsoft’s decision to enter the BI market with a
database rather than an end-user BI tool has provided a strong base from which to integrate end-
user products while offering performance capabilities unattainable in competing end-user BI tools
that relied on third-party, back-end server technology for performance.
In conjunction with its release of OLAP Services, Microsoft released the OLEDB for OLAP
specification that became a de facto multidimensional database access standard even before the
SQL Server 7 shipped. This standard evolved into the XML for Analysis (XML/A) standard, which
has been adopted by a wide variety of server and client tool vendors. There are now many client
and server products in the market that have adopted XML/A, which provides a robust platform
with a variety of choices for developers and end-users alike.
In 2000, Microsoft released SQL Server 2000. In this release, OLAP Services was renamed
Analysis Services. Its feature set was broadened, scalability was increased and data mining
capabilities were added to make Analysis Services ready for enterprise deployment.
In 2002, Microsoft released SQL Server Accelerator for BI (SSABI), a rapid application
development tool for the Microsoft technology stack. Unlike other rapid application development
products on the market at the time, SSABI generated the relational and multidimensional
schemas based on a metadata description rather than relying on a fixed schema tailored to an
application.
In early 2004, Microsoft released Reporting Services, a managed reporting solution. Reporting
Services is a component of SQL Server 2000 and completed Microsoft’s BI server offerings. With
Reporting Services, Microsoft introduced Report Definition Language (RDL), an XML grammar to
describe the layout and query information of a report. Reporting Services broke with the tradition
of monolithic reporting servers by separating report definition from report rendering. This
architecture opened Reporting Services to third party enhancements and application
development customization. Furthermore, Reporting Services reports can be automatically
generated from business processes and report output can be integrated seamlessly into custom
applications with a unique look and feel.
In 2005, Microsoft released SQL Server 2005, substantially re-architecting major components of
its BI offering.
• Integration Services, which replaced DTS, provides a better paradigm for managing ETL and
brings greater scalability.
• Analysis Services has a new data model (the Unified Data Model or “UDM”) that allows for a
more realistic business view of data with an improved approach to creating calculations and
managing key performance indicators (“KPIs”). Analysis Services also adds support for real
time or near real time querying without a dramatic degradation in query performance.
• Reporting Services adds Report Builder, an ad-hoc report creation and modification tool, and
a graphical MDX query builder. These features give developers and power users the
capability to generate reports without knowing SQL or MDX.
7 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
In SQL Server 2005, the development and management tools are based on Visual Studio 2005,
which gives the product line a greater integration with .NET, provides support for team
development, enhances debugging capabilities, and improves productivity by leveraging existing
skill sets among application developers and IT professionals. Further evidence of the integration
with Visual Studio 2005 is the introduction of reporting controls that can be embedded within
applications to easily embed SQL Server Reporting Services into applications.
Microsoft’s BI strategy of starting with the back-end analytical engine and moving forward to
support end-users has resulted in a broad BI offering that is architected to support a myriad of
both Microsoft and third-party developer and end-user tools.
An Evaluation of the BI Offering of Business Objects
At its core, Business Objects is a set of independent reporting, query, and analysis tools tied
together in a portal with a common environment for managing access and distribution. While each
of these tools does its job well, each is a separate tool servicing different user communities with
very little integration between the tools. The design environment for each tool is very different,
which reflects the differences in their evolution.
With the release of Business Object XI, Business Objects has made strides to unify its product
line. Building on Crystal Enterprise, Business Objects Enterprise forms the basis for its BI
offering, providing management and deployment of the reporting, query, and analysis tools.
InfoView is the portal that coordinates the access to documents from the tools. The core of the
tool set is Crystal Reports, Web Intelligence, and OLAP Intelligence. The Designer creates
Universes that are semantic layers that hide the complexity of SQL from users building queries.
The unifying factor in the Business Objects product line is the Universe. A Universe acts as a
layer between the data source and the query. Universes can be used with Crystal Reports and
Web Intelligence. However, a Universe is not applicable to OLAP Intelligence because an OLAP
data source has its own business metadata. When the Universe was first introduced, it was an
innovative concept and assisted non-technical users in creating relational queries. Universes
have not kept up with the needs of the market and add little to a BI infrastructure.
Business Objects does not provide client tools to manage the BI infrastructure and does not
address the back-end data needs of BI, such as:
• Aggregate management, which is an important aspect of a BI system given the current data
volumes.
• Real time BI, which is growing in importance and requires a different approach to data
acquisition than data warehouses or marts.
• Tools that assist an enterprise to manage its BI data, other than Acta (which is not widely
used).
Though there are application program interfaces (APIs) for most aspects of the Business Objects’
product line, these APIs do not really provide the type of tight integration required by enterprises
today. The only thing that has been integrated is a collection of reports and queries. A Business
Objects table, grid, or graph has a specific Business Objects look and feel that is usually different
than the user interface design of the rest of a BI application. This is not truly embedding BI within
an application, but rather it is wrapping a set of tools together within a BI application.
While Business Objects puts itself forward as a complete BI offering, its product line does not
cover the breadth and depth that a BI offering requires to meet the diverse needs of its BI
stakeholders. Client tools united together do not constitute a BI platform. Business Objects’ client
tools are targeted at executives and managers for performance management. Serving this
community is just an element of full BI platform. Client tools are just one piece of the software
8 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
stack required to create and manage a BI infrastructure. Without a mechanism to manage and
understand the back-end data, vertical integration of the BI stack is missing. This gap creates a
potential disconnect between the back-end and the front-end that lengthens development time
and can create problems.
Probably most limiting about Business Objects’ approach to BI is that its BI offering results in a
closed BI solution. Once an enterprise commits to Business Objects, the enterprise has severely
limited its extensibility options and there is little leverage of the existing infrastructure and skill
sets. Business Objects does not support a vigorous third party software market of tools that fills in
the gaps in its BI offering. Business Objects has not been adopted by application vendors as a BI
application deployment tool. The investment in Universes can only be used by Business Objects
tools: there is no leverage in using Business Objects, other than within its own tools.
An Evaluation of the BI Offering of Microsoft
Unlike Business Objects, Microsoft builds its BI offering from the analytical server up. This
provides a solid foundation for the many different BI roles within an enterprise. Building from the
data management up to the client tools gives an enterprise the ability to centrally manage BI and
leverage its use throughout the enterprise.
With the release of SQL Server 2005, Microsoft has created a backbone of BI server
components. In addition to its flagship relational database engine, Microsoft has the full
complement of required BI elements. Integration Services manages the flow of data in or out of a
BI system and across components of the BI system. Analysis Services provides analysis, metrics,
and key performance indicators, as well as real time analytics. Reporting Services provides
managed and ad hoc report authoring and delivery.
Microsoft has a unified and integrated view of metadata. Data source views describe the schema
of the source data. Report Builder puts the schema in context and allows power users to easily
navigation through the information and build ad hoc reports. The Unified Dimensional Model adds
the business perspective and supports metrics and key performance indicators. Any tool or
application can take advantage of the metadata to provide the relevant view of the information to
appropriate users.
The strong integration of SQL Server with Visual Studio gives the application developer a unified
environment for creating and maintaining BI applications. Sharepoint Portal Server provides a
comprehensive portal environment for BI within an enterprise. This is not just a BI portal, but a
single point of access for all information within an enterprise. Microsoft Office is also crucial.
Excel is the most popular BI client tool. With pivot tables and Excel Add-In for Analysis Services,
Excel becomes an integral part of a BI application. Office 12 will only add to this integration when
it is released.
The large array of partners that have aligned with Microsoft BI reveals the strength of its platform.
The use of open standards such as OLEDB, XML/A, and RDL gives the partner community a rich
set of capabilities that allow them to extend the platform easily. These partners provide:
• Analytical tools – such as ProClarity and Panorama
• Query tools – such as Business Objects and Cognos
• Excel add-ins – such as IntelligentApps and XLCubed
• Analytical applications - such as OutlookSoft and GEAC
• Programming components – such as eBlocks and ChartFX.
In addition to the foregoing, there are many features within the Microsoft platform that provide
support for moving BI into new uses. Proactive caching gives system support for real time BI. By
building support for real time BI right into the engine, Microsoft allows real time BI applications to
9 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
be built more easily and with greater reliability. The integration of SQL Server and the .NET
framework with Visual Studio 2005 allows embedded BI applications to leverage the skill set and
knowledge of the application developer without the worry of compatibility issues.
Summary
Business Objects provides a disparate collection of client tools with limited tool integration, little
vertical integration, and without support for open standards that discourages the leveraging of the
existing infrastructure and does not leverage existing developer skill sets. Microsoft provides an
integrated collection of server and client tools, full vertical integration, and full support for open
standards that leverages existing infrastructure and developer skill sets.
About Symmetry Corporation
Symmetry Corporation is a recognized expert in the design, development and implementation of
advanced business intelligence solutions for Fortune 1000 companies and software vendors. BI
is our sole focus and has been for nearly 20 years. A Microsoft Certified Gold Partner for BI,
Symmetry is a longstanding member of the Microsoft BI Partner Advisory Council and a key
contributor to the SQL Server Accelerator for BI (SSABI). Symmetry also created ADAPT, the
first database design methodology developed specifically for multidimensional database
applications, based on sound OLAP design principles. For more information, visit Symmetry’s
web site athttp://www.symcorp.com or call (415) 453-7966.
The Symmetry logo is a registered trademark and ADAPT is a trademark of Symmetry Corporation. Other company and
product names may be trademarks of the respective companies with which they are associated.
10 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
doc_264906793.pdf
Choosing a Business Intelligence (BI) offering is an important decision for an enterprise, one that will have a significant impact throughout the enterprise.
Evaluating Business Intelligence Offerings:
Business Objects and Microsoft
www.symcorp.com
August 2, 2005
Evaluating Business Intelligence Offerings:
Business Objects and Microsoft
Table of Contents
Introduction.................................................................................................................................... 3
What is Business Intelligence...................................................................................................... 3
Key Considerations in Deciding on a BI Offering ...................................................................... 4
Requirements of Business Users................................................................................................. 4
Requirements of Application Developers..................................................................................... 4
Requirements of IT Professionals................................................................................................ 5
Summary of Key Considerations ................................................................................................. 5
Comparing the BI Strategies of Business Objects and Microsoft............................................ 5
A Historical Overview of the BI Offerings of Business Objects.................................................... 6
A Historical Overview of the BI Offerings of Microsoft ................................................................. 7
An Evaluation of the BI Offering of Business Objects .............................................................. 8
An Evaluation of the BI Offering of Microsoft ............................................................................ 9
Summary ...................................................................................................................................... 10
About Symmetry Corporation .................................................................................................... 10
2 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings:
Business Objects and Microsoft
Introduction
Choosing a Business Intelligence (“BI”) offering is an important decision for an enterprise, one
that will have a significant impact throughout the enterprise. The choice of a BI offering will affect
people up and down the chain of command (senior management, analysts, and line managers)
and across functional areas (sales, finance, and operations). It will affect business users,
application developers, and IT professionals. Before deciding on a BI offering, business decision
makers must evaluate a company’s BI offering with respect to what it has to offer each of these
stakeholders. This paper will analyze the key considerations for deciding on a BI offering, review
the BI strategies of Microsoft and Business Objects, and then evaluate each company’s offerings
in light of these key considerations.
What is Business Intelligence
BI is a broad topic, covering many different functions (e.g., reporting and analysis) and
technologies (e.g., data warehouse, OLAP, portal). An examination of the literature shows many
varying definitions of BI. These definitions fall into two classes: a technical description of the
components that make up a BI solution and an explanation of the business purpose of BI. A good
definition of BI that encompasses both technical functionality and business purpose is the
following: BI is a broad category of applications and technologies for gathering, storing, analyzing,
and providing access to data to help enterprise users make better business decisions. To
effectively evaluate the offerings of BI vendors, how well an offering satisfies both sides of the
equation must be examined.
BI applications include the activities of decision support systems (DSS), query and reporting,
online analytical processing (OLAP), statistical analysis, forecasting, and data mining. Another
way of phrasing this is that BI applications take data that is generated by the operations of an
enterprise and translate that data into relevant and useful information for consumption by people
throughout the enterprise.
To deploy BI successfully throughout an enterprise requires a platform, not a hodgepodge of tools
and technologies. As such a BI platform must be able to address all the BI needs of an
enterprise. Gone are the days where BI was a point solution used by a limited set of highly
trained analysts. In today’s market, BI is pervasive throughout an enterprise and in many cases
mission critical. As such, a BI platform must be able to address both the breadth and depth of the
needs of an enterprise.
3 of 10
3 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
Key Considerations in Deciding on a BI Offering
Business decision makers must evaluate both the breadth and depth of a BI offering in order to
make an appropriate decision on a BI offering. The breadth of a BI offering is a measure of how
well the BI offering supports the different requirements of the BI stakeholders. The depth of a BI
offering is a measure of its vertical integration, i.e., how well the BI offering enables a business to
take raw data from a production system and transform it into relevant and useful information and
then deliver this information to business users in the proper context.
There are many steps required to generate consumable BI from raw data and there are various
BI stakeholders within an enterprise that have a vested interest in the process. IT professionals,
application developers, and business users (who are also the consumers of BI) all play a role in
the development of a BI solution. Business users define the business rules that determine how
the raw data must be transformed. Application developers develop the processes for acquiring,
consolidating, and presenting the raw data based on the business rules. IT professionals manage
the processes, ensure availability, and enforce security.
Requirements of Business Users
Different types of business users have different requirements of a BI offering.
• Analysts – Analysts support managers with performance management analysis. Analysts
require a powerful and interactive environment that allows them to create metrics and
navigate the data in an ad-hoc setting. This type of user requires tools for analytics, statistics,
predictive modeling, and advanced visualization.
• Managers – Managers at all levels need BI to assist them in making informed business
decisions. This type of business user requires a friendly query environment that also supports
the ability to generate ad hoc reports and delivery mechanisms that enable managers to
disseminate information up and down the chain of command.
• Operations workers – Operations workers use BI as part of solving a larger issue. For
example, as part of servicing a customer, a retail clerk might recommend other related
products to a customer. This type of worker requires BI that is embedded in a production
application, rather consuming BI as part of a BI application.
Business users of all types want to reduce their dependence on IT, but still have confidence in the
numbers, have advanced analytics, superior query performance, and access to timely information
in the format and delivery mechanism of their choice, whether through a portal, a spreadsheet, or
email. Satisfaction of these business user requirements enables BI to truly become mission
critical, fulfilling the promise of BI, and providing businesses with competitive advantage in the
global marketplace.
Requirements of Application Developers
Application developers must be able to develop the variety of BI application types required by
business users that are essential to enable enterprises to obtain competitive advantage in the
global marketplace. The range of capabilities that a BI offering needs to support is as varied as
the BI applications required by enterprise business users. BI applications such as sales analysis
need to be able to handle large data sets (terabyte) with very long lists of dimension members (in
the millions). Other BI applications must support complex calculations for the derivation of key
performance indicators or financial reporting modeling. Other BI applications merge BI analytics
with data collection for budgeting, planning, and forecasting. Still other BI applications require
very low data latencies for use in business activity monitoring applications to create real time BI.
The BI applications developer requires a BI offering that is capable of supporting this entire range
of BI applications.
In an enterprise today, BI is frequently embedded in business processes that support operations
workers and needs to be seamlessly integrated into existing applications, and then easily
extended as new BI needs are discovered. Application developers must be able to use existing
4 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
skill sets, reuse code and components, and leverage existing applications and infrastructure to be
able to meet the increasing need for BI that can be easily maintained and extended without
requiring new skill sets. A crucial factor for the productivity of application developers is having a
single development environment that allows them to work on all aspects of BI, from the data to
the analytics to the user interface using a single development language, and that supports team
development.
Further, application developers must be able to easily extend BI applications using third party
tools and technologies where necessary. These third party products must integrate existing BI
applications, rather than operate parallel to the platform. This requirement is an acceptance of the
fact that there is no perfect product in the market. A BI offering might fulfill most of an enterprise’s
needs, but there are always a few cases where requirements cannot be met with the existing
offering. Extensibility offers a safety valve in such cases.
Requirements of IT Professionals
IT professionals require a BI offering that enables them to deliver mission critical BI; namely a BI
solution that is highly available, reliable, and secure. IT professionals require a back-end solution
that is fault-tolerant and scalable; that supports change control and scriptable deployment; and
that enables them to leverage their existing resources and skill sets while building on the current
IT platform and infrastructure. Furthermore, IT professionals must be able to deliver real time or
near real time data to business users with minimal degradation in query performance.
Summary of Key Considerations
The scope of BI uses is large enough that trying to meet all of the needs discussed above with a
portfolio of stand-alone tools is inefficient, costly, and ultimately ineffective. Yet in order to provide
leverage, the BI offering must have a consistent manner of accomplishing tasks while integrating
with the existing infrastructure. This consistency will result in reduced training cost for business
users, application developers, and IT professionals; shorten development time for new
applications and enhancements; and result in greater acceptance by business users.
In summary a BI offering must:
? Support the end-to-end building of a BI application, from data source to the presentation
of information
? Encompass the diverse business user needs of analysts, managers and operations
workers
? Provide a solution for the varied application areas such as finance, sales, and production.
? Address the needs of various BI stakeholders in the enterprise: business users,
application developers, and IT professionals.
Comparing the BI Strategies of Business Objects and Microsoft
Business Objects and Microsoft have very different strategies for approaching the BI market,
which has resulted in very different BI offerings.
Business Objects started with a client tool for business users in 1990 and then pursued a
backwards integration strategy to provide a more complete BI platform for the IT professional and
application developer arena with its BI Application Foundation for analytical application
development. The strategy of Business Objects has been a strategy of acquisitions to evolve into
a broader back-end platform while still maintaining its focus on its original end-user audience.
Microsoft entered the BI market in 1997 with a high-performance analytical server offering and
pursued a forward integration strategy to provide a more complete BI platform based on an open
architecture supporting both Microsoft and third-party developer and end-user tools compatible
5 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
with its analytical server offering. Because Microsoft chose to enter the BI market 7 years after
Business Objects, the initial technology base for the Microsoft product reflects a more modern
architecture. The Microsoft BI architecture also reflects Microsoft’s strategic focus on a wide
audience of IT professionals, application developers, and business users.
A Historical Overview of the BI Offerings of Business Objects
Business Objects’ entry into the BI marketplace in 1990 was a client tool that displayed data from
relational databases in tables and graphs before such functionality was considered BI. Business
Objects differentiated itself in the marketplace by providing a semantic layer (called a “Universe”)
that insulated the business user from the physical database structure details. A Universe is
basically a way to rename columns to more appropriate business terms, create new calculated
columns, and hide the complexity of table joins.
In 1997, to meet the growing need of web based deployments, Business Objects introduced Web
Intelligence, a browser-based query tool. While the initial functionality of this tool was limited,
Web Intelligence is now the premier query tool within the Business Objects product line.
As the concept of BI took hold in the market, Universes were enhanced to represent dimensions
with hierarchies and measures. Also, the client side was enhanced to provide better BI-style
analysis, such as top x queries.
As Business Objects recognized the growth in OLAP multidimensional databases, Business
Objects purchased OLAP@Work to obtain both an Excel-based query tool and a vehicle for
access to Microsoft’s Analysis Services. OLAP@Work became the basis for BusinessQuery.
In 2001, Business Objects released Application Foundation, an analytical application framework
for jumpstarting analytical applications. Its Dashboard was initially part of the Application
Foundation and has since been split out as a separate offering. The Application Foundation
contains three analytical engines:
• Set analysis for segmentation of various business perspectives such as customer or
suppliers
• Predictive analysis for time series forecasting
• Statistical process control for Six Sigma and Total Quality Management.
Despite its name, the Application Foundation provides features more oriented to end-users than
IT professionals and application developers, which is a reflection of Business Objects’ traditional
orientation toward the end-user.
Next, Business Objects acquired Acta, a data integration vendor for Data Integrator, its extraction,
transformation, and loading (ETL) product. Data Integrator has failed to gain market traction and
has languished since its acquisition by Business Objects. In 2003, Business Objects acquired
Crystal Decisions, which gave Business Objects a managed reporting solution with Crystal
Enterprise and Crystal Reports, and an OLAP query and analysis tool in Crystal Analysis (which
is now known as OLAP Intelligence).
Business Objects has filled out its product line mostly through an acquisition strategy. While this
strategy has allowed Business Objects to rapidly expand its product line, this rapid expansion
strategy has come at a price. Business Objects has experienced difficulties in integrating
companies located in diverse geographical areas with different business structures and cultures,
and has not always been successful in this integration, as in the case of Acta. The acquisition of
diverse products has also provided distinct integration challenges. The most recent release
Business Object XI has repackaged the entire Business Objects suite of products. Before
Business Object XI, the Crystal and Business Objects product lines were separate and distinct.
With this recent release, Business Objects has brought common administration to the product
6 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
line, though the tools themselves remain discrete. Furthermore, Web Intelligence and Crystal
Reports still have different file formats.
A Historical Overview of the BI Offerings of Microsoft
Microsoft entered the BI marketplace in 1997 with the release of SQL Server 7. There were two
aspects to Microsoft’s initial foray into BI. The first aspect was OLAP Services, which was based
on technology that Microsoft acquired from Panorama Software. The second aspect was Data
Transformation Services (DTS), a data integration tool. OLAP Services was well-received
because it lowered the cost of entry for analytics and raised the bar for performance and
scalability in a multidimensional database. Microsoft’s decision to enter the BI market with a
database rather than an end-user BI tool has provided a strong base from which to integrate end-
user products while offering performance capabilities unattainable in competing end-user BI tools
that relied on third-party, back-end server technology for performance.
In conjunction with its release of OLAP Services, Microsoft released the OLEDB for OLAP
specification that became a de facto multidimensional database access standard even before the
SQL Server 7 shipped. This standard evolved into the XML for Analysis (XML/A) standard, which
has been adopted by a wide variety of server and client tool vendors. There are now many client
and server products in the market that have adopted XML/A, which provides a robust platform
with a variety of choices for developers and end-users alike.
In 2000, Microsoft released SQL Server 2000. In this release, OLAP Services was renamed
Analysis Services. Its feature set was broadened, scalability was increased and data mining
capabilities were added to make Analysis Services ready for enterprise deployment.
In 2002, Microsoft released SQL Server Accelerator for BI (SSABI), a rapid application
development tool for the Microsoft technology stack. Unlike other rapid application development
products on the market at the time, SSABI generated the relational and multidimensional
schemas based on a metadata description rather than relying on a fixed schema tailored to an
application.
In early 2004, Microsoft released Reporting Services, a managed reporting solution. Reporting
Services is a component of SQL Server 2000 and completed Microsoft’s BI server offerings. With
Reporting Services, Microsoft introduced Report Definition Language (RDL), an XML grammar to
describe the layout and query information of a report. Reporting Services broke with the tradition
of monolithic reporting servers by separating report definition from report rendering. This
architecture opened Reporting Services to third party enhancements and application
development customization. Furthermore, Reporting Services reports can be automatically
generated from business processes and report output can be integrated seamlessly into custom
applications with a unique look and feel.
In 2005, Microsoft released SQL Server 2005, substantially re-architecting major components of
its BI offering.
• Integration Services, which replaced DTS, provides a better paradigm for managing ETL and
brings greater scalability.
• Analysis Services has a new data model (the Unified Data Model or “UDM”) that allows for a
more realistic business view of data with an improved approach to creating calculations and
managing key performance indicators (“KPIs”). Analysis Services also adds support for real
time or near real time querying without a dramatic degradation in query performance.
• Reporting Services adds Report Builder, an ad-hoc report creation and modification tool, and
a graphical MDX query builder. These features give developers and power users the
capability to generate reports without knowing SQL or MDX.
7 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
In SQL Server 2005, the development and management tools are based on Visual Studio 2005,
which gives the product line a greater integration with .NET, provides support for team
development, enhances debugging capabilities, and improves productivity by leveraging existing
skill sets among application developers and IT professionals. Further evidence of the integration
with Visual Studio 2005 is the introduction of reporting controls that can be embedded within
applications to easily embed SQL Server Reporting Services into applications.
Microsoft’s BI strategy of starting with the back-end analytical engine and moving forward to
support end-users has resulted in a broad BI offering that is architected to support a myriad of
both Microsoft and third-party developer and end-user tools.
An Evaluation of the BI Offering of Business Objects
At its core, Business Objects is a set of independent reporting, query, and analysis tools tied
together in a portal with a common environment for managing access and distribution. While each
of these tools does its job well, each is a separate tool servicing different user communities with
very little integration between the tools. The design environment for each tool is very different,
which reflects the differences in their evolution.
With the release of Business Object XI, Business Objects has made strides to unify its product
line. Building on Crystal Enterprise, Business Objects Enterprise forms the basis for its BI
offering, providing management and deployment of the reporting, query, and analysis tools.
InfoView is the portal that coordinates the access to documents from the tools. The core of the
tool set is Crystal Reports, Web Intelligence, and OLAP Intelligence. The Designer creates
Universes that are semantic layers that hide the complexity of SQL from users building queries.
The unifying factor in the Business Objects product line is the Universe. A Universe acts as a
layer between the data source and the query. Universes can be used with Crystal Reports and
Web Intelligence. However, a Universe is not applicable to OLAP Intelligence because an OLAP
data source has its own business metadata. When the Universe was first introduced, it was an
innovative concept and assisted non-technical users in creating relational queries. Universes
have not kept up with the needs of the market and add little to a BI infrastructure.
Business Objects does not provide client tools to manage the BI infrastructure and does not
address the back-end data needs of BI, such as:
• Aggregate management, which is an important aspect of a BI system given the current data
volumes.
• Real time BI, which is growing in importance and requires a different approach to data
acquisition than data warehouses or marts.
• Tools that assist an enterprise to manage its BI data, other than Acta (which is not widely
used).
Though there are application program interfaces (APIs) for most aspects of the Business Objects’
product line, these APIs do not really provide the type of tight integration required by enterprises
today. The only thing that has been integrated is a collection of reports and queries. A Business
Objects table, grid, or graph has a specific Business Objects look and feel that is usually different
than the user interface design of the rest of a BI application. This is not truly embedding BI within
an application, but rather it is wrapping a set of tools together within a BI application.
While Business Objects puts itself forward as a complete BI offering, its product line does not
cover the breadth and depth that a BI offering requires to meet the diverse needs of its BI
stakeholders. Client tools united together do not constitute a BI platform. Business Objects’ client
tools are targeted at executives and managers for performance management. Serving this
community is just an element of full BI platform. Client tools are just one piece of the software
8 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
stack required to create and manage a BI infrastructure. Without a mechanism to manage and
understand the back-end data, vertical integration of the BI stack is missing. This gap creates a
potential disconnect between the back-end and the front-end that lengthens development time
and can create problems.
Probably most limiting about Business Objects’ approach to BI is that its BI offering results in a
closed BI solution. Once an enterprise commits to Business Objects, the enterprise has severely
limited its extensibility options and there is little leverage of the existing infrastructure and skill
sets. Business Objects does not support a vigorous third party software market of tools that fills in
the gaps in its BI offering. Business Objects has not been adopted by application vendors as a BI
application deployment tool. The investment in Universes can only be used by Business Objects
tools: there is no leverage in using Business Objects, other than within its own tools.
An Evaluation of the BI Offering of Microsoft
Unlike Business Objects, Microsoft builds its BI offering from the analytical server up. This
provides a solid foundation for the many different BI roles within an enterprise. Building from the
data management up to the client tools gives an enterprise the ability to centrally manage BI and
leverage its use throughout the enterprise.
With the release of SQL Server 2005, Microsoft has created a backbone of BI server
components. In addition to its flagship relational database engine, Microsoft has the full
complement of required BI elements. Integration Services manages the flow of data in or out of a
BI system and across components of the BI system. Analysis Services provides analysis, metrics,
and key performance indicators, as well as real time analytics. Reporting Services provides
managed and ad hoc report authoring and delivery.
Microsoft has a unified and integrated view of metadata. Data source views describe the schema
of the source data. Report Builder puts the schema in context and allows power users to easily
navigation through the information and build ad hoc reports. The Unified Dimensional Model adds
the business perspective and supports metrics and key performance indicators. Any tool or
application can take advantage of the metadata to provide the relevant view of the information to
appropriate users.
The strong integration of SQL Server with Visual Studio gives the application developer a unified
environment for creating and maintaining BI applications. Sharepoint Portal Server provides a
comprehensive portal environment for BI within an enterprise. This is not just a BI portal, but a
single point of access for all information within an enterprise. Microsoft Office is also crucial.
Excel is the most popular BI client tool. With pivot tables and Excel Add-In for Analysis Services,
Excel becomes an integral part of a BI application. Office 12 will only add to this integration when
it is released.
The large array of partners that have aligned with Microsoft BI reveals the strength of its platform.
The use of open standards such as OLEDB, XML/A, and RDL gives the partner community a rich
set of capabilities that allow them to extend the platform easily. These partners provide:
• Analytical tools – such as ProClarity and Panorama
• Query tools – such as Business Objects and Cognos
• Excel add-ins – such as IntelligentApps and XLCubed
• Analytical applications - such as OutlookSoft and GEAC
• Programming components – such as eBlocks and ChartFX.
In addition to the foregoing, there are many features within the Microsoft platform that provide
support for moving BI into new uses. Proactive caching gives system support for real time BI. By
building support for real time BI right into the engine, Microsoft allows real time BI applications to
9 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
Evaluating Business Intelligence Offerings
be built more easily and with greater reliability. The integration of SQL Server and the .NET
framework with Visual Studio 2005 allows embedded BI applications to leverage the skill set and
knowledge of the application developer without the worry of compatibility issues.
Summary
Business Objects provides a disparate collection of client tools with limited tool integration, little
vertical integration, and without support for open standards that discourages the leveraging of the
existing infrastructure and does not leverage existing developer skill sets. Microsoft provides an
integrated collection of server and client tools, full vertical integration, and full support for open
standards that leverages existing infrastructure and developer skill sets.
About Symmetry Corporation
Symmetry Corporation is a recognized expert in the design, development and implementation of
advanced business intelligence solutions for Fortune 1000 companies and software vendors. BI
is our sole focus and has been for nearly 20 years. A Microsoft Certified Gold Partner for BI,
Symmetry is a longstanding member of the Microsoft BI Partner Advisory Council and a key
contributor to the SQL Server Accelerator for BI (SSABI). Symmetry also created ADAPT, the
first database design methodology developed specifically for multidimensional database
applications, based on sound OLAP design principles. For more information, visit Symmetry’s
web site athttp://www.symcorp.com or call (415) 453-7966.
The Symmetry logo is a registered trademark and ADAPT is a trademark of Symmetry Corporation. Other company and
product names may be trademarks of the respective companies with which they are associated.
10 of 10
©Copyright 2005 Symmetry Corporation. All rights reserved. www.symcorp.com
doc_264906793.pdf