Business Intelligence and Analytics Fundamentals
Business Intelligence and Analytics Fundamentals
Business Intelligence and Analytics Fundamentals
and analytics
fundamentals
Helena Schwenk
www.ovum.com
Ovum provides clients with independent and objective analysis that enables them to
make better business and technology decisions.
With access to Ovum’s research and support from best-in-class analyst and consulting
teams, companies can turn analysis and insight into action. Our aim is to make our clients’
planning more effective, and to help them identify and assess relevant business
opportunities. We don’t just advise our clients: we collaborate with them to help them
exploit these opportunities and to turn them into business results. At the heart of our
approach is our mission to help, to be approachable, responsive and focused on your
business issues and to provide pragmatic and actionable advice and recommendations.
Ovum research is based on independently audited methodologies that ensure that our clients
can base decisions on rigorous and fact-based research, rather than on unqualified and
unjustified opinions. The research draws upon over 400,000 interviews a year with business
and technology, telecoms and sourcing decision-makers, giving Ovum and our clients
unparalleled insight not only into business requirements but also the technology that
organisations must support.
Business intelligence and
analytics fundamentals
SUMMARY
Impact
Business intelligence (BI) and analytic technologies are seeing rapid growth in today’s tight
economy, driven by fierce competitive pressures. Also fueling growth is the technologies’ ability
to improve decision-making, identify new business opportunities, maximize cost savings, and
detect inefficient business processes.
Ovum view
In order to survive and thrive in this competitive situation, organizations increasingly need to
shift the basis of differentiation to focus on the smarter and more efficient use of information,
thereby improving decision-making and creating sustainable competitive advantage. Hence, the
ability to extract, integrate, analyze, and interpret business information in a timely, proactive
way makes BI an increasingly important, influencing factor for organizations. We believe defining
a BI strategy will help shape and govern the success of a BI project by bringing greater business
and IT alignment and allow organizations to specify how timely, accurate, and relevant
information can support and drive core business initiatives.
At the same time a BI strategy may also prompt adoption of new BI suite technologies, lead to
new styles of deployment, and even create a process change that allows an organization to
become more flexible and agile and increase business performance at both strategic and
organizational levels. In short, a BI suite can increase the impact of employees in an
organization and the industries and sectors they represent.
Key messages
• BI is much more than a set of software tools. It describes a set of processes and technologies for
simplifying and enhancing the use of information within an organization. In particular, BI
provides a data foundation together with an associated analytic environment that can support,
underpin, and improve decision-making capabilities in organizations. For example, it allows
organizations to improve customer interactions, retain the best customers, maximize customer
satisfaction, improve customer profitability, provide business insights for strategic and
operational decision-making, negotiate more skillfully with suppliers, support greater
transparency into compliance efforts, and pinpoint operational inefficiencies crying out for cost
cutting.
Hence, the ability to extract, integrate, analyze, and interpret business information in a timely
and proactive way makes BI more of an imperative than a luxury for organizations in today’s
tight markets and economy.
• A BI suite comprises tools and infrastructure designed to bring actionable information to a broad
range of users. A BI suite consists of four key layers: a data management layer for sourcing,
integrating, and cleansing data; an analysis layer for deriving insight from that data; an
application layer for applying domain expertise and business logic to a particular business
problem; and a presentation layer for delivering actionable information and insight to users
regardless of context or location. To be effective, a BI suite needs to leverage components
across all layers and provide a flexible architecture that allows each component to be integrated
with the existing IT infrastructure to take advantage of existing investments
and facilitate faster deployment.
• Newer software deployment models are gaining traction. We are seeing significant changes in the
way BI and data warehousing systems are built and deployed. Most notable is the emergence of
alternative BI deployment options such as SaaS, purpose-built appliances, cloud computing, pre-
built analytic applications, and open source, which are starting to disrupt the traditionally
lengthy, expensive way data warehousing and BI systems were deployed in the past.
Organizations that want to increase speed to value and lower the cost of building and
maintaining their IT infrastructure should consider these alternative deployment modes.
However, they should be aware that these are still in the early stages of development and
implementation best practices have yet to mature.
Definition
BI and analytics are terms that refer to a variety of tools and technologies used to gather, store,
and analyze detailed transactional business data to support and aid organizational decision-
making. To understand the full scope and remit of BI, we believe its definition needs to cover
both processes and technology. In particular, BI includes:
BI may also be referred to as “business analytics,” a newer moniker that aims to emphasize the
proactive, forward-looking, and optimization elements of decision making.
We believe this new naming trend reflects the ongoing consolidation in the market where the
larger players have acquired or developed both BI and analytics technologies. We believe the
there are five key business trends driving the adoption and uptake of BIgrowth in the use and
adoption of analytic elements such as predictive modeling and data mining will help propel the
next phase of growth in the BI and analytics market as organizations attempt to model their
organizational data in a much deeper and more predictive sense.
BUSINESS IMPACT
Overview
The demands of business, competitive, and regulatory pressure in today’s economically
challenging environment are forcing organizations to become smarter and more efficient about
the use of their information. In particular, we believe technologies and solutions.
SELECTION CRITERIA
BI strategy comes before technology selection
Selecting the right BI software tool or analytic application requires planning and careful
consideration. However, before any detailed planning or IT conversation gets under way,
organizations need to devise and agree on a BI strategy that outlines a vision for how BI will
meet the needs and objectives of the organization. They also need to define a roadmap for
how the BI technology components of the BI program will fit together to meet these needs.
However, more often than not, organizations overlook this point and put the data and IT
architecture, rather than the needs of the business, at the top of the BI agenda.
We believe the best BI solutions are those that are closely aligned to the specific and unique
needs of the organization and use a BI strategy to meet this aim.
Primarily, a BI initiative needs to be a business-led engagement that is inclusive and encourages
broad participation from across the organization. Only when an organization has a clear
agreement of the business requirements can it start to plan and select its architecture and BI
suite components. Technology should be seen as an enabler of the overall business objectives by
providing relevant, accurate, and timely analysis of information that can help drive decisions.
However, there is another (complementary) way of categorizing BI end users – those who use BI
for strategic purposes versus those who use it at an operational level. Strategic BI is typically
characterized by historical analysis desi
designed to support long-term
term organizational goals by
looking at trends and patterns in the data and/or using multidimensional analysis to determine
the location or cause of a major problem. Operational BI is designed to support day-to-day
day
business decisions where
here BI is embedded in operational systems and processes. It works at a
more granular level by taking triggers from key business events and serving up context-relevant
context
information in role-based
based dashboards or interfaces. A more detailed breakdown of the different
dif
aspects of operational and strategic use is shown in Table 3.. Although BI usage thus far has
concentrated on strategic application, the rapid pace of business means organizations
increasingly need to provide support for decision
decision-making within daily operations. An important
consideration should be whether a BI suite supports all types of users and usage.
Figure 2 shows which evaluation dimensions to use when assessing the level of support provided
against an organization’s key requirements.
Components of a BI suite
A BI suite is designed to meet the full range of user needs and evolving business needs. It
provides a comprehensive suite of tools and infrastructure designed to bring actionable
information and greater business insight to a broad range of users, allowing them to access the
right information at the right time. An example of the layers found with a BI suite
te is shown in
Figure 3
Data management and data integration service
The data management layer is responsible for structuring and integrating the datain a format
that is accessible for analysis. The layer has two main functions.
Firstly, it handles the management and storage of data, metadata, and master data to ensure
that it is in a form that is optimized for information retrieval and analysis. In a BI suite, this is
often a data warehouse, although it can be a data store located closer to the operational data
sources (called an operational data store).
Secondly, the data integration platform provides a set of tools and services for profiling,
accessing, transforming, cleansing, and bringing together data, regardless of format, structure,
latency, or location.
The following tools are used as part of the data management and data integration
service layer:
• Data warehousing is a repository for storing relevant, consolidated, and cleansed data and
metadata that is optimized for the analysis and reporting needs of an organization or group of
users.
• Master data management (MDM) is responsible for ensuring the consistency, accuracy, control,
and synchronization of master data between operational and analytic systems.
• Metadata management provides a repository and associated tools for storing, integrating, and
analyzing metadata for the purposes of data governance, impact analysis, and data lineage.
• Data integration is responsible for the lifecycle and integration of data, whether it is integrated
by physically being moved from a source to a target data warehouse system, or whether it is
consolidated virtually within a federated single view.
• Data discovery/profiling is a process for discovering and examining the data available in a data
source and collecting statistics and information about that data to understand the location,
content, quality, and structure of source data.
• Data quality is used to standardize, validate, match, and cleanse data so it is fit for business
purposes.
Analysis service
This layer provides the analytical environment and exposes a range of services for processing
and analyzing information according to analysis needs and level of sophistication. The tools
provide both foundational BI capabilities, coupled with the deeper business analytics to help
organizations understand what is happening, why things are happening, and what needs to
happen in order to meet the needs of the business.
• Online analytic processing (OLAP) is a tool for analyzing key metrics interactively along multiple
dimensions for comparative reporting and analysis.
• Query tools extract and analyze business information for ad hoc reporting purposes.
• Predictive and advanced analytic tools use sophisticated data mining
algorithms, statistical models, and combinations of mathematical calculations to analyze current
and historical facts and make predictions about future events.
• Rule engines can be used to decide how data is processed and whether data is within an
acceptable range or has exceeded a threshold.
• In-memory BI tools enable faster processing and analysis of data sets by loading data into
memory rather than processing on disk. In some cases they eliminate the need to resort to
summarized data, pre-built OLAP cubes, or IT intensive database tuning.
• Simulation tools provide an environment for modeling and testing new ideas or for modeling
uncertainty. The tool supports a framework for comparing possible outcomes and helps clarify
the potential risks and rewards of pursuing actions.
Application service
This layer leverages the previous two layers to source and process the data, but also combines
its domain expertise, business logic, and predefined content (such as models, templates, and
reports) to address a particular business pain point, streamline a business process, or look for
new business opportunities. The application components can function independently as analytic
applications, but are often also embedded within the core transactional systems. Examples of
applications include:
• Enterprise performance management – provides capabilities for planning, budgeting,
consolidation, profitability, cost management, and the alignment of strategic plans with
operational performance by monitoring metrics and KPIs.
• Sales analysis – provides support for activities such as product sales and trends, product
affinity, commission reporting, and salesforce performance.
• Marketing analysis – provides support for activities such as campaign management, marketing
channel efficiency, demographic analysis, loyalty programs, and upselling and cross-selling
analysis.
• Supply chain analysis – provides capabilities that include supplier intelligence, demand-driven
forecasting, and quality control analysis.
Information access and presentation services
This layer is responsible for locating, accessing, and presenting information to the user. The key
technologies include:
• Search – in a BI context, is used to locate and retrieve BI reports and analysis. However, the
technology is increasingly being used to reveal previously unforeseen relationships in data and
documents, and makes it easier to find both unstructured and structured relevant information
within a single interface.
• Publisher – is responsible for outputting business information in a format for the screen or print
using multiple file types, such as Adobe PDF, HTML, Microsoft Excel, PowerPoint, Rich Text
Format, or plain text.
• Alerts – provide automatic notification to an end user, usually by email, if a trigger or threshold
set by the business has been reached or breached.
• Dashboard – organizes, aggregates, and presents information in a visually appealing way,
typically using KPIs tailored to a particular user audience, such as executives, line-of-business
managers, or call center managers.
• Mobile BI – provides remote access to select BI information via smartphones and mobile
devices.
• BI widgets/gadgets – present dynamically updated and personalized BI information on a user’s
desktop.
SOLUTION MATURITY
A consolidating market landscape
We estimate BI license, maintenance, and service revenues to be $7.1 billion in 2010, growing
to $9.1 billion in 2014, which represents a 6% CAGR over the period. (See Figure 4.) At the same
time, the BI software market has undergone some significant changes in the past five years,
resulting in a shake-out of vendors and the emergence of new technologies and deployment
models. To date, the market has consolidated around a handful of mega-suppliers that now
dominate the BI market landscape. This spate of consolidation was mainly driven by the
market’s solid growth rate and the potential for value creation in areas such as the provision of
integrated BI suites, integration and bundling with other parts of the software stack, and
opportunities for cross-selling.
Currently, an eclectic mix of vendors dominates the market for BI software, including:
• the big four – SAP, Oracle, IBM, and Microsoft
• independent BI and data warehousing vendors such as SAS, Teradata, MicroStrategy, Qliktech,
Information Builders, and Kalido
• smaller disruptive players such as open source companies (Jaspersoft, Actuate, and Pentaho),
appliance vendors (Netezza, ParAccel, and Greenplum), and cloud BI providers (Birst,
Swiftknowledge, and Oco).
The big four BI vendors now account for the lion’s share of BI software market revenues.
However, this consolidation does not mean an end to the independent BI market in which a
vendor’s main product focus is BI. Even though the gap between the big four BI vendors and the
other providers is widening, opportunities remain for the latter to shine and differentiate by
focusing on specialist
cialist and vertical niches.
For example, SAS, Teradata, and MicroStrategy have built up vertical expertise across a number
of sectors, such as retail, telecommunications, and manufacturing – which sometimes entails
partnering with one another – to offer ttailored and industry-specific
specific BI suites. Similarly,
increasing consolidation within the BI market and hardware and software commodization have
provided opportunities for smaller niche and specialist BI providers, such as open source, cloud,
and appliance vendors.
RECOMMENDATIONS
Recommendations for enterprises
In addition to best practice, enterprises are encouraged to take into account the following
technology considerations.
APPENDIX
Glossary
• Analytics. Often used interchangeably with BI, analytics refers to a higher plane of analysis,
typically forward-looking, driven by statistical trending/forecasting, predictive analysis, and
data-mining techniques.
• Analytic application. A software application designed to facilitate faster deployment by providing
prebuilt content such as extraction routines, models, templates, and reports to address a
particular business pain point, streamline a business process, or look for new business
opportunities.
• Appliance. An integrated, preconfigured set of servers, storage, operating system, database
system, and application software that is purpose built for a specific function – such as BI
analysis or data warehousing.
• Business intelligence (BI). Includes a set of tools, processes, and applications that gather,
integrate, query, analyze, and report on information to help organizations make better business
decisions.
• Cloud computing. A computing model for delivering commercial IT hardware or software
services to companies over the Internet.
• Data warehousing. A centrally held and integrated repository of standardized, clean, consistent
information sourced from various operational systems, which is structured in a way to make it
readily amenable to BI query, analysis, and reporting.
• Event stream processing (ESP). Uses techniques such as complex pattern detection, event
correlation, abstraction, and hierarchies to identifying meaningful events within multiple streams
of event data. ESP enables applications such as algorithmic trading, fraud detection and location
based services.
• Enterprise performance management (EPM). Consists of a set of processes and applications to
help organizations optimize their performance by linking business strategy to operational
execution using a methodology and framework for linking and analyzing key performance
indicators against business processes.
• Latency. A time delay, such as between recording an event or transaction and acting upon it.
• Master data management (MDM). A combination of tools, processes, and people to establish an
ongoing competency in delivering a single source of truth for key data used within an
organization.
• Metadata. Information that describes or defines another piece of data. An XML schema is an
example of metadata, as it describes the structure of an XML document.
• Open source. An approach to the design, development, and distribution of software that offers
developers access to the software’s source code.
• Online analytical processing (OLAP). Provides multidimensional views of segments of enterprise
data and supports drill-down from summary level to detail.
• Software-as-a-service (SaaS). A model of software deployment whereby a provider licenses an
application to customers for use as service on demand over the Internet. SaaS software
providers either host the application on their own web servers or download the application to a
consumer device.
• Web services. An architecture in which software is delivered as a set of webbased services,
which can be called from any application, without regard to the underlying platform or operating
system.