Archives ||About Us || Advertise || Feedback || Subscribe-
Issue of July 2003 
 Home > Cover Story
 Print Friendly Page ||  Email this story

Cover Story: Business Intelligence

A strategy for intelligence

BI offers a number of benefits to enterprises. But how does a company know it's time to use BI. And does it need a separate BI team and an enterprise-wide BI strategy? Answers to these and more. by Soutiman Das Gupta

Business Intelligence (BI) tools allow companies to automate its functions of analysis, strategy, and forecasting to make better business decisions.

Companies generate raw data from various sources and touch-points within the organization, and from external entities like suppliers and distributors. And this raw data is stored as 'information' in OLTP systems like databases, ERPs, SCMs and CRMs. This information can be stored in a central repository like a warehouse, to help extract knowledge. A BI solution can then draw 'intelligence' from the warehouse.

“You know you have to use BI when the cost of coordination outgrows the cost of BI tools and centralization in the warehouse” Anup Bagchi, Head Retail Strategy & New Product Group, at ICICI Bank

Why do companies need BI?

"Looking at it broadly, enterprises need to derive intelligence for two main purposes. First, for performing analyses to help make decisions. The analyses can help identify sales trends, and provide alerts about important customers and complaints. Second, to help predict future customer behavior and market demand," explained Arun Rangaraju, Solution Manager, CRM/BI, of SAP India.

"Enterprises need to refine data into meaningful information so that business users can analyze it. Users tend to look at information in different perspectives. The information has to be thus presented in the
right context," added Balaji Jagannathan, National Manager, Information Management, of Computer Associates.

Survive without BI?

Can an organization survive without using BI tools? The answer is, yes. An organization can collect data from multiple sources and create the usual reports like monthly sales, profits, number of new customers, and inventory levels.

But for a large organization like a bank, insurance company, and teleco, which has close to a million customers, is it worth the effort to collate the sheets and reports? It's difficult to track patterns and behavior changes, peg growing segments, and understand market potential. Decision-makers also need cross-tab information to create strategies, analyses, and forecasts.

Moreover, as the size of the company grows and it's geographical spread increases, the nature of the reports becomes lengthy and complex. Simply running 'eyeballs' over them will not provide the macro picture.

Time for BI

"You know you have to use BI when the cost of coordination outgrows the cost of BI tools and centralization in the warehouse," explains Anup Bagchi, Head Retail Strategy & New Product Group, at ICICI Bank.

The business heads see that there is an explosive growth of data across nationwide systems. Reports are generated at various stages of the business at various departments and locations, and the business and product heads find it difficult to extract the relevant information needed to make business strategies and forecasts.

Mahesh Ramakrishnan, CTO of Reveleus, which is an i-flex business, points out three indications for an enterprise that says it's time to use a BI tool.

  • A business realizes it needs BI when there are large isolated islands of data. It takes a lot of time and effort to put information together.
  • Typically, this organization will have an army of people collating data from various sources, and providing information to those who need it. This is because no one can really source it from one place.
  • These complex scenarios cause a lot of pain since a lot of time is taken to acquire and publish information. And very often, important business decisions are made based on an approximation.

The enterprise perspective and relevance

Companies have always used technology that manages and stores enterprise data. BI is the technology that will manage the information, which has been provided by the data.

ICICI Bank has 5 million retail customers, 1.1 million credit card customers, and numerous other customers for other products like loans and bonds. If a bank of this size is able to identify an emerging need from a small fraction of its customers, it amounts to a lot of revenue.

Anup Bagchi of ICICI Bank says that 25 percent of the bank's asset business grew over the last 12 months through cross-selling. The asset business includes auto loans, home loans, and credit cards.

"Airlines claim a three to eight percent revenue increase by using yield management systems," said M.S.V. Rao, Director, Department of Information Technology, Air India.

Standard Chartered Bank (SCB) used OnLine Transaction Processing (OLTP) systems, which captured transaction-oriented applications. It answered the financial queries and generated reports at a broad portfolio level, which included total earnings, debt situation, interest income, cost, fee income, and profits.

"The system was reliable but provided little scope for in-depth customer analysis," said Sedjwick John Joseph, Head-Business Intelligence Unit, SCB.

This was the critical point at which the bank decided it needed to analyze the huge volumes of data captured by its OLTP systems. The idea was to search for crucial nuggets of information from the vast amounts of transactional data to get the right information, to the right executive, and at the right time.

BI teams in companies

"BI tools are used by companies to increase business and gain competitive advantage. Since this activity is very important to an organization, it's a good idea for a company to have a separate BI unit. This unit will analyze the results thrown up by the BI tools and fulfil the information requirements of departments across the organization. This is an emerging trend in the BFSI and telecom industries in India," said Bill Gibson, CTO-Asia Pacific, SAS International.

"Even with the BI data a CIO cannot take a marketing decision randomly without consulting the various division heads from Sales & Marketing, Finance, Product development, Operations, and Customer service. For example in a sales and marketing scenario, BI helps to understand customer needs and responses to new market opportunities. The CIO with his sales team can gauge the effect of pricing and promotions, target customer segments more accurately, analyze buying patterns to take advantage of cross-sell opportunities, and develop
true real-time one-to-one marketing activities," said Tarun Malik, Product Manager-Business Tools Division, Microsoft India.

Ready teams

Indian organizations like the RBI, BPL Mobile, ICICI Bank, and SCB have teams dedicated to the study of analysis of information and the ways in which it can generate value to the organization. All these companies have benefited from their respective teams.

However, Tariq Farooqui, Country Manager, JD Edwards India has a different view on this. "The results of BI are for the direct consumption of personnel in the functional teams. It should enable the business/functional personnel to analyze the information themselves. The analytical information can be used to find out information like the most profitable customer on which products are performing the best vis-à-vis competition," he explained.

BI enterprise strategies

A very important aspect an enterprise should understand when it comes to BI is that, BI is a business project aimed at business users for business benefits. BI is not essentially an IT affair.

Data warehousing and relevant tools enable the use of BI. Business users should formulate the parameters for transforming the data in a warehouse, and an IT person should code it. The BI team at the enterprise should then formulate a strategy to use the findings from the BI tools effectively.

SCB used BI tools to discover that a significant proportion of its business came from the upwardly mobile Indian woman. And business was likely to grow substantially from this section in the coming months. This gave the bank's BI team an impetus to launch a product, which catered to the needs of this segment.

SCB launched DIVA, an international credit card targeted at the Indian woman and bundled it with several features. It includes discounts and zero interest rates on categories like jewellery, cosmetics, apparel, consumer durables, leather products, and mobile phones.

BI must be treated like a business initiative. It's function, use, and benefits must percolate down to all levels in an organization. The deliverables must be tracked and finally it's all about action.

An enterprise must act wholeheartedly and quickly with the help of the information from BI systems. The information might lose relevance in a matter of weeks, or even days.

Alok Bansal, Solution Architect, Baan Info Systems India talks about BI strategy in enterprises. "The right strategy for an enterprise will be determined by many factors. They are industry business drivers, the company's perceived strengths, weaknesses, opportunities and threats. Whatever the final choice, the goal will be the same for all enterprises, namely to create, increase, and maintain an elusive quality known as competitive advantage."

Soutiman Das Gupta can be reached at

Telecom Industry Critical Success Factors
Why BI goes bad

Organizations may have spent a lot of money to buy BI solutions and hire consultants. And it may have invested in quality hardware and software to build a warehouse.

But even then, some may not be satisfied with the performance. They may feel that the tools do not show the desired results, and the business is not able to derive the expected benefits. What could have gone wrong?

“The BI tool and the data warehouse should be trained and re-trained dynamically so that it's ready to provide intelligence for tomorrow” Deepak Verma, Senior Vice President and Chief Operating Officer, BPL mobile

Here are a few possible reasons:

  • The BI project and activity is construed as an IT project. The team running BI may not be a business team. It may be a team of IT personnel with very less representation from the business users' side.
  • The data warehouse and BI tools, and technology are implemented for technology's sake, and not with specific benefits in mind.
  • The implementation phase of a BI project is neglected. Implementation of a warehouse and BI requires blood and sweat. It needs an enormous amount of detailing and customization.
  • There are not enough inputs from various departments in an organization. The BI project needs big organizational involvement. It cannot run by itself.
  • The IT personnel do not understand the organization's business aspects. To make BI successful, IT personnel must understand business over a period of time.
  • The ETL (Extract Transform Load) script has been made entirely by IT personnel. The business user must formulate/create the ETL parameters, and IT personnel must script it.

Deepak Verma, Senior Vice President and Chief Operating Officer, BPL mobile added to this:

  • The vendor was not asked for the 'proof of concept.' This can be explained with an example.Findings from a BI tool should provide actionable and relevant data. The BI tool shouldn't churn out data that is voluminous and irrelevant. It should list the most probable cases for up-sell, cross-sell, or turnover. It should help the operating teams with a focused list that can help them act in a cost effective manner.
  • The BI is not evolved at the company. Markets and environments change. To keep pace, a company must re-train its knowledge. The BI tool and the data warehouse should be trained and re-trained dynamically so that it's ready to provide intelligence for tomorrow.
Use internal and external data intelligently

M.S.V. Rao, Director, Department of Information Technology, Air India

Business Intelligence (BI) covers the functions of gathering, processing, and analyzing large volumes of data from the internal system and external sources. And this is possible because BI uses sophisticated tools of analyzing and forecasting at a speed, which helps an enterprise take real-time decisions to achieve organizational objectives.

Organizational objectives

Enterprises use technology to achieve the following basic orgnizational objectives :

  • Reduce costs
  • Improve productivity
  • Improve the product
  • Improve customer service
  • Improve revenues

BI derived by analyzing the internal and existing data can contribute significantly to achieve these objectives. And the use of enterprise applications like ERP and technologies like Intranet have facilitated the use of BI.

“Using Yield management, airlines forecast the class-wise demand at departure of a particular flight”

BI helps

The increase of digitization enables enterprises acquire phenomenally high quantities of data. This gives them an opportunity to convert this data to information, and knowledge. BI is pedagogically recommended to help improve revenue and customer service. At this stage, enterprises need to cross boundaries of their organizations, acquire and analyze data from the external world, including customers, competitors, and environment.

Strategic information

BI allows enterprises to exploit large quantities of information by analyzing it to find behavior patterns of their customers and competitors. This can significantly help an organization suitably modify its plans in various aspects of business like production, distribution, pricing, and capacity planning. Online access to such information can help decision-making and bring dynamic changes to help improve a company's bottom line.

BI in airlines

BI is very relevant to a service organization and the airline business is no exception. Due to high use of IT in an airline, these companies can use BI to their advantage. The very nature of an airline business makes it necessary to exploit BI to survive in the highly competitive and price-sensitive market. Two major applications of BI have taken root among airlines. One is of yield Management (YM) and the other of Market Intelligence (MI).

Using YM, airlines forecast the class-wise demand at departure of a particular flight. This is done by analyzing the booking and cancellation pattern on that flight using data of the past 12 to 18 months on similar flights. This coupled with the fares offered in different classes will optimize the number of seats offered in a particular fare class on any day for a future flight. Airlines claim a three to eight percent revenue increase by using YM systems.

The second well established application in airlines is the processing of MI Data Tapes (MIDTs) supplied by Global Distribution Systems (GDSs). GDSs connect thousands of travel agents to hundreds of airlines and therefore have millions of transactions stored in their systems. Airlines buy this data at very high costs and analyze it to understand the distributors (agents) and the marketing segmentation.

Stiff competition and similarity of operational characteristics between airlines have seen airlines identify customer value as the major discerning factor. Customer value is best understood by analyzing the numerous CRM transactions and effecting changes immediately.

At Air India

Air India has implemented a Revenue Management System in 2001-2003 and has experienced increased revenues by being able to adjust the fare levels dynamically. We will also obtain MIDT data and analyze it to derive MI. On a smaller scale, we have implemented a data mart to store uplifted coupons (tickets sold) to help the marketing team analyze the pattern of our actual flight load and revenues.

Advice for companies
Individual data marts can help different business units analyze the data and take effective steps. But the ideal growth path is to develop an enterprise data warehouse.

This can help collect data about customers, competitors, and the environment through various channels and make this information available to all the units of an organization. Such availability of data aided with technologies like OLTP, statistical analysis, and data mining will give the full benefit of enterprise-wise exploitation of BI.

Business Intelligence market in India

In the recent Infrastructure Strategies 2003 survey conducted jointly by Network Magazine and IMRB, Indian companies in different verticals were asked what their top IT priorities in 2003-2004 were.

32 percent of telecom and ITES companies, and 24 percent of Banking and Financial Services Industry (BFSI) companies said that Knowledge Management (KM) was the area of highest priority in 2003-2004. This opens up a lot of opportunity for the introduction of BI in these companies.

A Frost and Sullivan Business Intelligence report (2001-2008) says India is still in the adoption phase of enterprise applications like ERP, SCM, and BI. BI, which is a much more mature application, is likely to be deployed once the primary applications are in place. BI's effectiveness is hampered if the information infrastructure of the organization is not in place.

However, the Indian market is likely to witness high growth rate during the period 2002-04. The demand is likely to be fuelled by large and medium-sized enterprises, and MNCs. A positive trend is that many organizations are going in for data warehousing, data mining, OLAP on transactional data, and some data mart suites to address the needs of specific business units or departments.

Anup Bagchi on creating enterprise strategy and customer wealth

Head Retail Strategy & New Product Group at ICICI Bank Ltd

Organisations have since time immemorial been using information to make faster and better decisions. The genesis really has been with management wanting answers related to business performance. Management may ask questions of the ‘why’ variety. For example, ‘why’ have dipped this quarter as compared to the last?

Banks broadly use Business Intelligence (BI) to drive strategic organic growth through leveraging information assets to optimise costs, strengthen customer profitability, and new product development. In order to increase customer profitability, a bank can cross-sell a multitude of owned financial products like bank bonds, loans, and investment solutions as well as a mix of third-party products like stocks and Government bonds.

BI enables enterprise strategies tailored towards up selling, campaign management, channel and value migration. The quest is to identify segments of customers that show common investment patterns, and plan new products and services.

Channel Migration

Using BI tools, an analysis of transaction patterns could give important insights into consumer behaviour. Customer behavior, payment pattern, and transactions can be used to assess customer profiles.

For example a bank can analyze customer segments who predominantly use branch banking channels to carry out transactions which, could ordinarily also be carried out at low cost electronic delivery mechanisms. Strategy could then be formulated to migrate such segments through a process of consumer education and motivation to ATMs, phone banking, and Internet banking.

Value Management

Analytics assist banks to leverage knowledge to manage customer value. Banks need to know aspects like the segment of customers whose deposits grow over a period of time, and the segment whose balances decrease or stagnate over a period of time. Based on this understanding banks can take steps to arrest downward migration, accelerate upward migration, and move a stagnated user profile to an upward migration path.

An organized source of data and the use of BI tools can help a bank achieve these objectives.

A warehouse

In Core banking applications the focus is to facilitate daily operations and manage transactional integrity. A data warehouses on the other hand not only is an integrated source of historical data of customers across products and channels but is also geared towards the running of complex queries.

Using BI tools, one can access relevant information and present it to business analysts or decision-
makers in a form, which is intuitive, user friendly and actionable.

The need for integrated, automated BI tools is accentuated when the cost of coordination outgrows the cost of technology and centralization in the warehouse. For example, a bank may have to build MIS teams of 5 people for each product or service. A product-centric approach of this nature will mean that with every new product launch, there will be linear additions in people doing MIS. In addition to this, maintaining information consistency and arriving at a ‘single version of the truth’ becomes a real challenge.

A simple thumb rule for gauging the impact of BI on the organisation is to analyse marginal value i.e. what has been done with BI, and what could not have been done without using these tools. We can attribute a large part of the benefits accrued by cross-selling, identifying transactions behavior, channel-migration, and value management to the use of BI and a warehouse.

- <Back to Top>-  

© Copyright 2001: Indian Express Newspapers (Bombay) Limited (Mumbai, India). All rights reserved throughout the world.
This entire site is compiled in Mumbai by the Business Publications Division (BPD) of the Indian Express Newspapers (Bombay) Limited. Site managed by BPD.