Business Intelligence

Business Intelligence for the masses

Cost effectiveness, and reliable and measurable results, are driving BI implementation in customer-focussed verticals such as retail, telecom and manufacturing. Today, BI tools and their benefits are part of the business mainstream. Kumar Dawada reports.

Knowledge is the antidote to fear
—Ralph Waldo Emerson

For centuries, prosperous and successful businesses have depended on intuition and the analytical skills of elite management personnel. Timing has always been critical--the right decision at the right place and at the right time can be sustained with insight, foresight and solid performance backups. The IT revolution has, to some extent, reduced business insight to mathematical formulae that have been standardised and made available as Business Intelligence (BI) systems that dredge facts to support the best possible business decision from an organisation's data stores. In practice, businesses deploy these solutions for competition analysis, predicting trends, customer needs, buying patterns and behaviour based on existing data.

Internal Intelligence

IDBI Bank has developed its Business Intelligence in-house. Sneha Khanna analyses the deployment at one of India's largest private banks

Banks cannot live without Business Intelligence as they rely heavily on technology. IDBI Bank has its own internally developed framework for BI.

INTERNAL MANAGEMENT

Sanjay Sharma

Says Sanjay Sharma, Corporate Head, Technology, IDBI, “We do not use a standard tool for BI. We built our own tools for customer profitability and account profitability, looking at data from different dimensions and creating product line related information.”

IDBI preferred this option as they have the information on their core system. Sharma says, “Instead of extracting everything from the core system and then dumping it somewhere and extrapolating from it, it is easy to work within the organisation with a set of people and so we created the model internally. We also used the external perspective of some experts and small organisations where a set of people worked onsite to help us develop these models.”

A PHASED IMPLEMENTATION

IDBI implemented BI in a phased manner. Firstly they looked at the larger picture of their database to study the customer segments that they wanted to target. Then they analysed and found out who the profitable customers were.

After detailed analysis of their customers, IDBI Bank created an analytical and behavioural model. It took them around six to eight months to create the BI framework.

SELF INTEGRATION

Running an internal BI framework as it does, the bank rarely faces the problem of data integration. Says Sharma, “For CRM, everything has been developed in-house. CRM and BI are running on the same data and so we don’t face the problem of data integration as there is no third-party involvement.”

That changes when a tie-up arises. IDBI has a solution for data integration issues in such instances. Sharma explains, “We have many tie-ups with companies such as BPCL, and more recently, Indian Airlines. In these cases, there is online connectivity and backup is done using the ERP system for online interfacing with XML as a message-based format and SSL for security. In these tie-ups there is a predefined validation process.”

At IDBI there is a specific format that needs to be validated and uploaded into the core system for any exchange of data. So there is no inconsistency.

COUNTING ON BENEFITS

IDBI has benefited from day one, enjoying success apart from cost benefits. Sharma explains, “When there is a subset of data in a large database, BI helps in targeting the mailers and phone calls that are part of any product campaign. The conversion rate for direct sales and product relationship has been almost thirty to forty percent.”

In the past IDBI has launched many products. Consider for instance, when IDBI launched the Card-to-Card Money Transfer. Recalls Sharma, “When we realised that there are lot of people using drafts and other methods to transfer funds, particularly for individuals studying abroad or travelling, we launched Card-to-Card Money Transfer where one is able to transfer from one visa debit card to another non-IDBI debit card.”

Another strategy which IDBI made and followed was when they observed that every Diwali and New Year the number of credits in their customers’ accounts increase as many companies directly credit accounts, some giving cash or cheques. Sharma reveals, “We launched gift cards not only targeted at individuals but also at companies that can give an incentive or bonus equivalent to the value of the card, which customers can use in shops or at ATMs.” IDBI observed that the customers with a balance of Rs 50,000 to Rs 1,00,000 and with 10 transactions a month might say yes to wealth management and offered them the service.

OBVIOUS PARAMETERS

Sharma believes that for any organisation to evaluate its success for BI there are some basic parameters, such as cost reduction over time, additional revenue generated, quantum of sales growth and so on. Customers are getting the most out of BI through value-added service.

If a customer has a balance in his account which has not been used for a long time the bank offers advisory services suggesting mutual funds that he might be interested in.

With business booming, the bank is now contemplating buying a packaged product. Sharma reveals, “Now we are planning to go in for a product instead of managing all our BI needs internally, as our information store is growing.”

Bonding BI with banking

A bank such as HDFC, with over three million customers spread across the country, needs to have a vast Business Intelligence framework to keep up with and analyse its customers' information. Let us examine how they do it.

Munish Mittal

A bank such as HDFC, with over three million customers spread across the country, needs to have a vast Business Intelligence framework to keep up with their customers’ information and analyse it. Let us examine how they do it.

Business Intelligence tools are the banking sector’s lifeline, helping to consolidate the vast volumes of data, and target the right segment for the right scheme. Success helps in reducing costs tremendously.

WHEN THE NEED ARISES

HDFC Bank started using BI in 2001 when their customer base expanded to over three million. Munish Mittal, Vice President, Information Technology, HDFC Bank explains, “Once we acquired critical mass with around three to four million customers in 2001, we had to integrate customer data into a central customer data repository to understand and review all services the customer was subscribing to, as well as customer profitability, to devise a customer-centric framework.”

Within six to nine months after the BI implementation, HDFC felt a tremendous difference in:

• Analysing customer behaviour and interaction with the bank.
• Devising a customer contact strategy and implementing suitable channels such as direct banking.
• Analysing customer preferences for transaction channels.
• Identifying personalised needs based on relationship value, and devising cross-sell strategies to arrive at the basic mission, which is to be a one-stop shop.
• Devising a suitable product strategy.
• Refining new products

THE 36-DEGREE VIEW

HDFC began implementation by first creating a 360-degree view of customers’ transactions, profitability, product acquisition, product holding, loan repayment, ticket sizing and transactional trends of usage, which helped them to offer many new things such as electronic bill cards.

Next, HDFC did data warehousing by analysing campaign management using a modelling engine by Unica on top for studying customer behaviour. It took around eight to nine months for HDFC to integrate data warehousing across product segments in retail banking.

BONDING THEM TOGETHER

Bonding BI to customers and suppliers was another issue that HDFC had to deal with and managed successfully. “For supply chain relationship management we have implemented solutions in the area of corporate banking, where the payment between the automobile dealer and manufacturer as well as the supplier and the same manufacturer can be executed in real time on the Internet,” reveals Mittal.

HDFC uses various Business Objects technology based end-user reporting and personal OLAP tools to study the behaviour of supplier, dealer and manufacturer transactions over the Internet to generate trends, devise pricing strategy and create innovative use of product offerings.

CLEANING UP

For cleaning up data, HDFC has in-built health check-ups and for Extract Transform Load (ETL) they have database administrators. Mittal explains, “For ETL we have database administrators who work with us to optimise ETL performance. We use innovative and state-of-the-art storage technology to derive maximum performance of IO (input/output) processing.”

For data integration, HDFC has health checks, data standardisation, and data enrichment as a part of their daily/weekly/monthly data warehousing processing. HDFC is currently implementing a data quality and householding solution from SAS for enhancement of its BI infrastructure.

—Sneha Khanna

Bajaj Auto: the turn-around move

Faced with a stagnating scooter market and increased competition from local and MNC players, Bajaj Auto went for a major restructuring and streamlining operation. It found a major ally in its enterprise portal and BI solution. It helped Bajaj Auto’s phenomenal turnaround and reinstated it as one of the two largest manufacturers of two-wheelers. Kumar Dawada reports

“The intelligent thinker sees the invisible, feels the intangible, and achieves the impossible.”
— Anon

In the late 1990s, Bajaj Auto, the leading scooter manufacturer, faced an identity crisis. There was a shift from scooters to motorcycles, whose designs were sleeker and more fuel-efficient. Urban youth as well as rural markets were patronising the motorcycle market, effectively ignoring scooters. It was a do-or-die situation for Bajaj Auto. It had to quickly adapt to the evolving market or be rendered obsolete.

Bajaj Auto went for a major organisational overhaul and restructuring. Different channels were created. The marketing network was improved. Product quality and innovation was given priority; the supplier database, cost conversion ratio and all other aspects of manufacturing and finance were improved.

PROJECT 110 PERCENT

In 2000, the company’s business objective was code-named Project 110 percent. The goal was to boost the company’s topline by 100 percent and the bottom-line by 10 percent. Bajaj Auto aimed at reducing its purchasing cost while increasing retail sales, dealer profits and customer satisfaction. IT structure became an important element in the overall business strategy.

Since 2002, Bajaj Auto is using the mySAP enterprise suite, which includes ERP, customer relationship management (CRM), supply chain management (SCM), and product lifecycle management (PLM), and is based on a transaction-processing model. Hence, SAP-BIW (Business Warehouse) was added as an extension to the mySAP enterprise suite to provide BI functionality and to ensure tight integration between the transactions, the system and the BI.

Anil Khopkar, GM, MIS, Bajaj Auto feels that the SAP BI solution is best suited for the manufacturing as well as the automotive sector. “The reason for deploying SAP BI solution was that SAP also builds and provides India-specific solutions. It has a development lab in Bangalore and ensures strong product support. There is also proper integration with previous products,” says Khopkar.

Khopkar calls the BI initiative a necessity. “The enterprise suite and the transaction system were taking care of day-to-day business. However, the BI deployment was required to meet corporate goals and objectives. Every individual had targets to meet. The BI initiative was necessary to monitor and measure target achievements. BI has enabled the organisation to move beyond streamlined business transactions and focus on performance management,” stresses Khopkar.

THE CORRECT VERSION

Even when the business setup and processes are exceptionally good, the challenge faced by an organisation is to ensure that the data available is a single irrefutable version of the truth. The data placed in the BI solution must be able to reflect the actual business transaction and measure it against the actual performance. If the data is unable to correlate the weekly or monthly business transactions or represent the business activities then it loses the confidence of decision-makers. “Our main concern was to ensure that the data available on annual, monthly and even daily basis should result in action orientation on the frontline and facilitate overnight decisions. What we needed was traceability of the data transaction so that people would start believing in the information provided by the IT system and take decisions accordingly,” recalls Khopkar.

Bajaj Auto decided to focus on its core competency, manufacturing, instead of IT intricacies. It partnered with IBM and then with SAP to take care of technical details including data cleaning, readying the data sources and overall care of data integrity. “The content was ours. The technical and IT aspects were taken care of by our support partners,” confides Khopkar.

“The BI initiative provided the company with tremendous data visibility especially for retail and market sales. It also improved quality control and cost structure. Visibility and availability of proper data makes people work better because it shows them the exact status and galvanises them into action. We make the information available at the proper level so that the concerned persons are able to relate to their business target and make the best of it,” elaborates Khopkar.

MEASURING EVERYTHING

The BI deployment is used to measure all business functions, goals and targets including marketing topline, market shares, customer satisfaction index and dealer performance index. The company also has a business scorecard model and a shareholder value system in place. Since there has been a single SAP system for many years, there is tight integration and no compatibility issues.

Khopkar believes that unless you can measure something in business by performancxe it cannot be achieved. BI has facilitated achieving business objectives. The achievements are measured, monitored and made action-oriented.

To ensure freshness of ideas and infusion of new technology, Bajaj Auto constitutes a core IT team with senior (level two) business people who represent the various business functions. They spend two to three years and then go back to their respective departments. “The new person coming in their place introduces fresh ideas, approach and technology and the new implementations are not challenged or blocked by old mindsets formed by investing too much time in the IT department,” says Khopkar.

Khopkar feels that the intangible RoI is most important in a major IT deployment. He explains, “The performance measurement enables our people in their daily operation and performance. We are now focusing on improving overall performance. We have set our sights on achieving excellence in all areas including production, manufacturing and marketing. We are also focusing on product innovations and are getting recognition in the form of excellence awards.”

Shopper’s Stop: closer to customers with BI

Shopper’s Stop wanted to make shopping a delight for their customers. They also wanted to understand their customers’ needs and provide better products while reducing costs and driving revenues. BI helped them achieve their goals.

Unnikrishnan T M

The retail sector is getting more competitive locally and is under siege from global retail giants such as Wal-Mart. The Indian retail industry, which initially lagged behind in adapting new technologies, has quickly realised that strategic use of technology alone can ensure their long-term survival. They have also pinpointed the areas where BI can be applied specifically to their industry.

RETAIL INTELLIGENCE

India is a very complex retail market. Retailers not only have to contend with geographical, cultural and linguistic diversity, but manage business complicated by tax structures which vary from state to state. A successful retailer needs not only the right products at the right place and the right time but must also have the right price to be acceptable to the customer. In addition, the number of products managed is very high.

Retailers are highly dependent on three types of Intelligence. First is merchandise intelligence. This includes everything related to merchandise, from product inventory to price and marketing. The second is customer intelligence, starting from knowing the customers’ buying patterns to ensuring customer loyalty. The third intelligence is operational intelligence. This ensures cost management and includes reducing retail shrinkage due to theft by employees, shoplifting by customers or vendor fraud. It also covers labour optimisation and proper selection of the shopping mall site.

PERSONALISED SHOPPING

Shopper’s Stop had a clear business objective. It wanted the customer to have a personalised shopping experience. It wanted tailor-made promotions and schemes unique to the buying patterns of specific customers. “For this we needed to understand how, when, where and in what combination the customer buys merchandise,” recalls Unnikrishnan TM, Customer Care Associate and CTO, Solutions and Technology, Shopper’s Stop.

It was also necessary to improve the decision-making skills of the organisation, manage costs, increase revenue and provide better products to the customers. It decided to use a BI solution from Business Objects as well as the SAS analytical application. It uses the Business Objects solution for reporting and some advanced analysis. “We use a combination of Business Objects and the SAS solution for trend analysis, promotion management, and customer behaviour, segmentation, buying basket analysis, profitability and lifecycle analysis. We also use it to understand the effectiveness of marketing efforts,” says Unnikrishnan. “The BI solutions were chosen for their ready availability as well as optimal cost at that point of time,” he adds.

CLEANSING THE DATA

Before any IT initiative, the business requirements as well as the process to be supported and workflow to be facilitated are carefully evaluated. This evaluation process ensures that the business processes are aligned with IT before deploying any IT solution including BI. “We use extensive qualitative and quantitative evaluation criteria. Every solution is graded on the criteria and vendors are expected to demonstrate their products’ capabilities on all the criteria,” emphasises Unnikrishnan.

As BI analytics depend critically on data integrity and reliable data resources, Shopper’s Stop have models for collecting, updating, changing and deleting data through the business processes. “We ensure that the business process supports the data lifecycle and has the necessary checks and controls to validate the data being entered,” says Unnikrishnan.

“The data which is transferred from online transaction processing systems (OLTP) has to be cleansed. As and when required, transformation is carried out and then this data is loaded onto the data warehouse. But first, a data mapping exercise that maps company-specific data to a retail industry standard data model should be completed. In some cases, a company may build their own proprietary data model. The choice may vary based on various factors,” elaborates Unnikrishnan.

He adds, “Our current BI applications are fairly compatible with the existing ERP, SCM and CRM applications. Even if compatibility issues arise, enough Enterprise Application Integration (EAI) tools and middleware applications are available to resolve the issue and harvest the data into the data warehouse. ETL tools too are easily available to extract data, and the transformed data can be easily loaded on the data mart or data warehouse.”

BI solutions are currently implemented to measure revenue parameters, customer enrolments, profitability of a customer base or segment and total revenue earned through the lifecycle of a customer. BI is also used to measure the merchandise mix.

“Business Objects is used for many business applications, business processes and functions. We also use BI for our CRM activities as well as for other areas of merchandising, inventory and sales as required,” stresses Unnikrishnan.

RICH DIVIDENDS

The RoI depends on the application for which the BI is used. BI implementations in some areas have paid rich dividends. “Birthday mailers are sent out to our loyal customers and the response to this initiative is phenomenal. We also use BI during the festival season to increase traffic to the stores. Here too the results have been encouraging,” says Unnikrishnan.

Shopper’s Stop has used various predictive methods and models to find out which of its customers are likely to buy again and how much. This has helped create a focused campaign and incentives for its high-value customer base. The recent diamond mine festival was a big hit with high-value customers as was its iPod giveaway scheme.

The organisation uses the information and analysis collected from the BI tools to help it study the demographic details of the target markets and position merchandise in retail store locations.

FUTURE OF RETAIL BI

BI will help the retail sector in understanding the customer. This insight will be used to make proper business decisions pertaining to various areas. In BI’s future is a fusion of intelligence with business processes.

—Kumar Dawada

Home loans and BI

With a network of 51 branches and 106 service stations in India, DHFL has a large network, a wide customer base, and a comprehensive product portfolio. The company uses BI to consolidate data and create reports that give it a deeper understanding of its business. Sneha Khanna reports

There are many big national and international players today in the housing finance industry. There is competition for space and market share. Free and fair access to the knowledge base, and appropriate business information are the key factors that help a company stand out in a service-based industry. Making information available in a snapshot is the strategy that Dewan Housing Finance Limited (DFHL) has been following.

FUNDAMENTAL NEED

For a company providing financial assistance, Business Intelligence (BI) is a backbone which forms the very basis of existence. Says Satish Kotian, Head, IT, DFHL, “Due to the nature of global competition, there is always pressure on profit margins and hence the cost of acquiring business, cost of delivery and delivery method are key factors in the success of any organisation.” Another reason which made DHFL choose BI was that this information was being used by various functional departments of the company as well as disseminated to various agencies as per their requirements. Says Kotian, “Crystal Reports captures the data in a processed and semi-processed form, and gives users the flexibility of customising business reports.”

To deploy Crystal Reports at DHFL, a core team was formulated for gathering the complete set of business and process information required for MIS. Another team was formed to evaluate the products available in the market on the basis of their capability to process data, security, portability, reliability, speed and deployment including scalability and future usability.

INSTANT RETURNS

Having deployed Crystal Reports approximately three months ago, DHFL is now in a better position for customer segmentation based on income range, saving potential, employment opportunities, loan size and geographical spread. Kotian says, “Business data is collected at every instance across the organisation, all over the country, and Crystal Reports gives us a ‘Business snapshot’ at one go. DHFL has been able to get a grip on recovery and defaulter analysis.”

After applying the BI tools, the company was able to analyse its collection data by area, group and industry.

INTEGRATION ON THE WAY

After the data is collected and processed, the next step is data integration. DHFL is validating old data, gathering and updating missing information and then migrating the data so that it is compatible with the data generated by its ERP and CRM systems and ready to use for BI.

The Home Loan Application is the company’s core application, where it tries to completely remove any form of data redundancy. Data appropriateness is automatically checked and verified as a part of the process validation.

CLEAN UP THE BAGGAGE

After data integration is data cleanup for which DHFL totally relies on the BI setup as a reporting tool. Says Kotian, “In a service industry like ours, data cleaning is the most important aspect. BI has helped us to identify real critical parameters which significantly impact our business. Immediate detection and intervention is possible through the analysis of such reports.”

DHFL hopes to complete the entire data transfer process of cleaning and transferring data in 6 to 12 months.

BENCHMARKING BI

For benchmarking, according to Sanjay Deshmukh, Business Development Director, Asia South, Business Objects, “Business Objects’ BI application can help Dewan Housing Finance identify trends in its business.

With this data, DHFL can decide which customers are likely defaulters and which customers can be cultivated further to cross-sell other products.”

The team can now spend less time accessing information, and more time on qualitative analysis. Users are able to analyse available information better.

Reliance Infocomm: Saying hello with BI

Using BI from the company’s inception has helped Reliance Infocomm (RIC) face market saturation and stiff competition, manage rapid growth, control the enormous data flow between systems, and achieve its business goals.

Rajiv Gupta

When Reliance Infocomm launched its telecom service, it had to face not only market saturation and stiff competition from established players but other major challenges as well. It had to address business opportunities such as offering products and services.

These opportunities included products not already available at competitive rates, customer acquisition, customer retention and up-sell. These were identified as key growth drivers and RIC felt that Business Intelligence (BI) was the critical business driver.

SETTING BUSINESS OBJECTIVES

To begin with, RIC had well-set business objectives for the BI solution. This included increasing competitive edge by understanding the customer better, improving product positioning, providing better customer service, making optimal use of IT resources, better and effective campaign management, getting maximum benefits from cross-sell, up-sell opportunities as well as other available business opportunities. It also included taking care of churn management, revenue management and controlling credit exposure.

A CUSTOMISED SOLUTION

Rajiv Gupta, Head, Decision Support System, Reliance Infocomm says that the BI solution at RIC is customised and consists of components such as Extraction, Transformation and Loading (ETL) tools, database, pre-formatted or canned reports and On Line Analytical Processing (OLAP) tools.

“We had discussions with several consultants including HP and IBM, and evolved a RIC-specific BI solution by choosing appropriate tools and vendors for the BI components. The primary reason for a customised solution was that it gave us the flexibility of choosing best-of-the-breed solutions for each component. This approach helped us leverage the available tools and products,” says Gupta.

“RIC’s key requirement is to have a scalable and open architecture which provides an unparalleled response to business requirements in an efficient, accurate and timely manner. It is also necessary to source the data from multiple systems, transform it based on business rules and store them in a repository at granular level to ensure that each transaction of the customer is captured and available for analysis,” says Gupta.

Other criteria for choosing the BI solution included flexible infrastructure for BI and analytics, strong administration capabilities, fast access and split-second response. “The customised BI solution was implemented after ensuring that it fully satisfied all key requirements, provided easy integration with available tools, products, components and domain knowledge, while providing a strong BI experience,” reveals Gupta.

In RIC, the BI solution was implemented right in the beginning. This ensured that it became the only system to provide information to all users spread across a large geographical area. “As this system has grown with the organisation and is built based on business inputs, it has become a part of RIC’s everyday life. To ensure data integrity we have standardised some business and IT processes. This includes business requirement gathering, application development, data retention policy and processes, quality and testing. We have also developed an enterprise data model across all objects and have common development standards and policies,” explains Gupta.

Gupta says, “RIC is already using the BI solution for customer lifecycle management, sales monitoring, channel profitability, collections management, network optimisation, product development and optimisation, and customer service.”

INTEGRATE WITH ALMOST ANYTHING

“RIC’s BI solution has the ability to integrate with almost all types of source and target systems. This has also allowed us to automate even the interfaces between systems. It is scalable to accommodate increases in source and target systems, record volumes and the number of users,” elaborates Gupta.

He adds that the BI solution also provides information at all levels. The top level is provided with dashboards for Key Performance Indicators (KPI) and exception reports. The middle management is provided with trends and analysis. The field staff benefits from operation and actionable information at the line item level. The BI solution is tightly integrated with all kinds of source systems. It sources information from Provisioning, CRM, Billing, Mediation, and ERP, and has the ability to write back to them.

RETURN ON BI

Gupta feels that RIC has benefited immensely from the investment made on the BI solution. More than 2,000 users currently use it across geographical locations very effectively. In time, the solution will be rolled out to more business areas and more users.

In the near future, RIC plans to update its BI solution to include predictive analysis, corporate performance management, dash-board and scorecard management, strategic enterprise management including business planning and simulation, and information workflow management.

Gupta feels that the next generation of BI solutions will not only show patterns from past data but also plan for the future in real-time. Predictive analysis will become more powerful and it will soon be possible to analyse live data.

—Kumar Dawada