The Hindu Business Line : In the cross-hair

Business intelligence is all about first locating targets before hitting them. And datawarehousing tools help a company do just that — locate and tap vital data buried in various departments and operations that can impact a company's profitability. It's a question of zooming in.

VIVEK Dayal is the owner of a retail store in Mumbai. Wouldn't he love to know what items get sold in a hurry at the beginning of the month and what in the latter part of the month? Imagine if nappies went out in a chunk in the first 10 days and then again in another chunk in the last 10. He needs to get them only every 15 days instead of buying a whole lot at the beginning of the month. A case of just-in-time inventory, based on the knowledge of how the sales pan out.

Cut to Biswarup Sengupta, the regional sales manager at a shirt-maker's. Wouldn't he love to know the status of his sales on a weekly basis instead of through the usual monthly reports? And that every time a good sales rep — Anup Menon — took 10 days off, shirt sales dropped 2 per cent? Sure, we all would love to know things that make us better at our jobs or, at the least, that make it easier for us at work. The IT industry calls this business intelligence, BI for short. And how does such intelligence come about? Clearly, from the past data that a company collects during the course of its functioning. This data, when collated and slotted, becomes information. When used to better one's functioning, this information is transformed into intelligence. For collating and categorising data, you need what is called the datawarehouse (DW). Thus came about the business of BI & DW which is translating into a huge opportunity for the IT industry. This week, eWorld takes a look at the industry that has become mainstream in the last four years or so. Interestingly, the IT industry's past efforts at digitisation at companies worldwide have led to a situation that provides more opportunities for the BI segment.

Enter datamarts

Recounting the evolution of DW, Vivek Bhasin, General Manager (BI & DW), Wipro Technologies, says "The actual practice of datawarehousing has been around longer than is generally thought. But it has become mainstream in the last four years. Front-end players in the market have helped the business user to publish reliable information that they so badly needed to run their operations." In other words, it led to a microstrategy, resulting in the formation of small datamarts distributed right across an organisation. "Getting these tools available in the market helped to make available solutions quicker. But it also led to issues. Issues that included solutions that helped particular departments, such as marketing, sales and finance, but doing little else for the top guy overseeing all these departments. So users were burdened with large silos or stand-alone systems that sometimes led to duplication of efforts by various departments or did not allow free flow of information between departments."

Says Bhasin, "In some instances we find between 10 to 50 independent silos that do not talk to each other." So there is a resource duplication as well as a technical duplication. This was bad in some aspects. But it wasn't necessarily the wrong way to start.

Says Kiran Cavale, Practice Director, BI & DW Practice, Satyam Computer Services, "For customers, it was easier and cheaper to build one data mart at a time. A decision to build a datawarehouse depends on the tools you use, the kind of data you have and the volume of data you handle. At the time, a datawarehouse might not have made sense." But once customers realise the value of a DW and know they need it, they typically graduate to having a BI strategy and a road map, Bhasin says. According to him, "We help define a user's requirements. Say, 5 to 10 applications over the next 2 to 3 years." After identifying the technology, Bhasin helps the user formulate a strategy that reduces total cost of ownership and at one stroke centralises the information. Now made available to all authorised to access this information and benefit from it, the warehouse actually impacts profitability for the better. This means information flows in from one central repository of data, thus maximising return on investment.

DW gaining acceptance

"When we said it is moving mainstream we mean that not only are Fortune 200 companies accepting it, but Fortune 2000 companies are also looking at such technologies," he says. Says N. Veeraraghavan, Practice Director, Data Warehousing Practice, Cognizant Technology Solutions, "There are five clear benefits to the average Chief Information Officer: cost savings, improved operational efficiency, better customer service (where applicable), faster decision-making and meeting core business demands." Veeraraghavan is quick to point out where BI could have helped but was not deployed. "The recent strategic scandals show that there was no business intelligence built into the financial reporting system. A clear lack of information is evident." Veeraraghavan is very excited when he talks of savings for the customer. He says, "If a user department makes a change in a single column on a data sheet, it has to inform the IT department. It typically takes 26 days and 3-4 people to make this change and revert to the user department. Now, with analytic tools, it is possible to make this change in 4-5 minutes. Business activity monitoring is a crucial application now — where companies have an almost-NASA like environment with data showing up red alerts or green signals depending on the results of the activity monitored." Bhasin says business intelligence is being expanded to assimilating operational data stores. Required information is taken out from the data stores instead of datawarehouses.

Key applications

Key applications engendered by BI include supply chain intelligence and understanding the customer better. In the former case, says Bhasin, "manufacturing and retail companies want efficient supply chains. So integrated strategy is the talk of the hour to facilitate global procurements and ensure the quality of goods produced or acquired." For example, if a customer wants to centralise global procurement, it is because that would help him negotiate better with his vendors. So costs go down significantly.

"Enhancing the Web channel is another application gaining importance. When a company's customers buy over the Web, knowing the customer demographics, past buying habits or past experiences in purchasing becomes crucial. BI steps in here." Agrees Veeraraghavan. "An integrated approach to the whole exercise of digitisation is the talk of the hour. ERP was the group we used to work frequently with earlier. Now, the CRM front-end team is who we work with often. Most requests for proposals are for customer-centric views. This is the trend we expect to see in 2003-04." The verticals most active in the DW space are banking and financial services, manufacturing, healthcare and governance. BI outsourcing is gaining momentum. Such outsourcing helps tie in the customer to the vendor. According to Bhasin, "We totally manage the systems. Even if we acquire the systems and not the assets, it still becomes a five-year contract." That's revenue on an on-going basis.

Employee numbers

Wipro has grown from about 100 employees for the BI & DW business four years ago to about 375 now. CTS set up the DW practice in 1997. CTS, says Veeraraghavan, "was able to hit the ground running thanks to the projects we had done for Dun & Bradstreet, our parent, which was essentially in the data vending business." Satyam Computer Services set up the DW shop in 1997 and has grown to 325 from 60 employees in about three years. How does Bhasin see the roadmap for his division? "Four years ago, talent in the Asia-Pacific region in this area was limited. That is not so now. Further, there are four levels of operation: maintain, integrate, architect, consult. Within a couple of years of starting off, we were into consulting. In this game, there is quick graduation from technical services to chief strategist." How do Indian vendors fare in the market? Bhasin feels, "The top global brands have been good at consulting but not good in implementation. Indian vendors have the advantage of offshore-onsite mix. So our customers are telling tool vendors to partner not only with pure consultants but also with implementation players." In the case of CTS, it came out with a framework for its clients in 1998. Veeraraghavan says, "We still use this framework. This year, our objective is to provide the integrated analysis framework that will form the basis for clients to architect their BI solution. Clients find this useful particularly in fraud detection and score carding." Finally, eWorld couldn't help asking this question: in this time of slowdowns and cost cutting, didn't budgets for BI get affected? Bhasin says, "If a company has to compete, information is critical. How does it do this at lowest cost? So, strategy is important. So is the choice of applications. Finally, return on investment is monitored on an ongoing basis in a BI project. If you don't have these systems, you don't save money. So, there is no choice to not having such a system."

Veeraraghavan admits, though, that sales cycles are certainly going up. Also, the increase in the number of vendors makes it difficult to choose whom to go with. Satyam Computer Services recently participated in a Datawarehouse Institute Conference 2002 in the US. Says Cavale, "There is no expected increase in dollar spending till end of 2003. But there is more and more offshoring happening. In other words, there is more expected out of lesser expenditure. The consolation for vendors is that datawarehousing has seen a mere 3 per cent cut in spending compared to CRM (38 per cent cut) ERP (9 per cent) and supply chain management 48 per cent." Satyam has an interesting perspective on the level of services required for a DW operation. Says Cavale, "The ratio of revenues from services to product licences is about 1:1 for datawarehousing as compared to 5:1 for ERP." Apparently, with demands for greater return on investment, DW and BI saw a peak in interest from interested CIOs. Says Cavale, "two years for a datawarehouse enterprise is too long. So, shorter sales cycles with lower spends was the order of the day. Digital cockpits and the like were other reasons why cuts in spending are lower in DW. After all, you need to invest in DW in order to save costs," he says, echoing Bhasin's view.

Offshore model

Says Cavale, "In the last six months, there has been definite improvement in the scenario for spending, but there are no concrete indications yet. But what is happening is the use of the offshore model as the trump card. Earlier the ratio of onsite to offshore was 75:25. Now that is reversing. So, datawarehousing and quality consultants have been brought offshore. This saves the customer some five times the onsite cost. In some cases we have also done the data modelling offshore. For some strategic studies, we have done that here for a project." He says that in terms of geography, the technology is predominantly US-led, with Europe contributing far lower than its US cousin. "If the total BI services revenue was $18.3 billion, the US contributes 53 per cent to it.

RTE

How about the real-time enterprise? If data helps make decisions faster, real-time data — through Internet technologies — should help make decisions even quicker. However, feels Veeraraghavan, that time has not yet arrived. He says, "No one wants it unless there is a crisis. In the case of fraud detection, RTE tools would help. But not generally."

Pie size

Wipro Technologies' revenues from BI & DW constitute about 5 per cent of the Enterprise Solutions business, which in turn contributes 59 per cent to the company's topline. For CTS, about 7 per cent of the company's topline comes from datawarehousing.

Cross-selling

How does an Indian vendor sell to a DW customer? Do cold-calls work? That is, can you simply walk in to a customer and hard-sell BI/DW? Or do existing customers for other businesses of the Indian vendor express a need for a datawarehouse? Says Cavale, "We do cross-sell. So, the datawarehousing arm has also brought in new clients. This helps other specialties within Satyam to pitch their wares to these customers.The reverse also happens." It is learnt that in 40 per cent of the cases, the enterprise applications vendor is the same as the datawarehouse vendor.

Obstacles to a good BI roadmap

BI, which is a mature application, is likely to be deployed once the primary applications are in place. The effectiveness of an application such as BI is hampered if the information infrastructure of the organisation is not in place. Sanjay Deshmukh, Business Development, Director, South Asia, Business Objects, says that non-availability of data on key subject areas for complete business analysis is the next big challenge. "For example, to understand the exact profile of a profitable customer a bank needs to capture the customer demographic information. For this, banks are looking at implementing CRM applications. All BI application software can be successfully implemented if the BI initiatives are driven by business users, not the IT department, he says.

Business users in new economy organisations

Banking, Finance, Telecom are driving BI initiative in organisations. But in old-economy companies and Government institutes, the business users lack the initiative for driving a Business Intelligence campaign. This is a key challenge for all the BI software application companies. Gourish Hosangady, CEO and Managing Director, SAS India Pvt Ltd, feels the issue of lack of data integrity is paramount. "BI demands reliable data for analysis which is found wanting in most enterprises. "There is a perception of high costs and returns on investment for BI implementation. This is despite the fact that BI deployments are known to deliver returns anywhere between 50- 200 per cent returns within 3-6 months of deployment. But, it is only a matter of time before the BI market in India shows significant growth,'' he says.

Market numbers

Looking at the global market size of BI, the market size in 1999 was $36.3 billion and the BI/DW industry is growing at 45 per cent plus, year on year and the total estimated market is close to $148 billion by year 2003. According to Gartner projections, BI/DW has come of age. By 2002, 50 per cent of IT service providers would incorporate the operational value of BI functionality into their go-to-market strategies. By 2003, BI functionality would be one of the most heavily weighted criteria enterprises consider when evaluating a potential IT solution. And these enterprises would treat BI expenditure as a competitive necessity and an indispensable cost of doing business. By 2004, 50 per cent of Global 2000 companies would have a corporate wide BI strategy and 2005 would witness BI being used by 70 per cent of corporate end users in midsize and large enterprises.

Through 2006, enterprises that implement BI initiatives using a BI framework would experience a return of at least two to three times greater than those who do not would. According to IDC, the BI segment is currently worth $1.2 billion, where 43.5 percent comes from Australia and New Zealand, while South East Asia, India and Korea contribute 35.3 percent. Greater China constitutes the remainder of the pie. The APAC market as per IDC will experience a compounded annual growth rate of 23 percent to touch $ 3.3 billion in 2006.

The market in India is expected to grow at CAGR of 30-40 per cent for the next 3 years. SAS India is a market leader at 22.5 per cent market share and 70 per cent top of mind recall as per Frost & Sullivan.

According to Frost & Sullivan, in 2005, revenues are expected to touch $30.4 million. The same study also reveals that the revenues will grow at a CAGR of 33.9 percent between 2002 and 2008. Compared to this, Gartner predicts the BI market to grow to $18.1 billion by the year 2006 from $11.3 billion in 2002, with a CAGR of 11.7 per cent. Among the worldwide markets, North America continues to be the top market with a potential of $15.3 billion (CAGR of 9.9 percent), followed by European markets with $9.6 billion (CAGR of 11.9 per cent). The other major markets include the Asia-Pacific with $1.56 billion and Japan with $1.42 billion by the year 2006 with a CAGR of 24 per cent and 12.6 per cent, respectively. "The financial year is not over for 2002, however, at this stage we feel comfortable that both China and India can reach the high double to triple digit growth in 2002. Based on current estimates on the business plans for 2003, we would expect a continuation of this trend in 2003,'' Hosangady of SAS India says.

Bhasin of Wipro, looking at the domestic market, says the BI market is expected to generate revenues of $10.7 million by the end of calendar year 2002. This number is expected to move further to touch a figure of $30.4 million by the end of 2005, according to a recent study conducted by Frost & Sullivan on business intelligence in India during 2001-2008. These revenues are expected to grow at a CAGR of 33.9 per cent over the forecast period.

The India factor

N. Veeraraghavan, Practice Director, Data Warehousing Practice, Cognizant, says compared to the US and European business intelligence markets, the Indian market is miniscule.

The challenges the market faces include lack of historical customer data and track record, low IT maturity, lack of product understanding, bottlenecks with integration, and ambiguity related to return on investment. A Frost and Sullivan report says the market restraints include low product awareness, high implementation costs, managing organisational culture, lack of positive testimonials and managing technology complexities. BI software needs to extract, assimilate, analyse information from diverse systems to derive meaningful information. Due to unclean data in a large proportion of enterprises in India, integration, customisation and upgradation of technology are regarded as a major hindrance.

The four stages

How does a datawarehouse get built? There are four stages to this. First, the vendor comes in and spends time understanding user requirements. Here a business analyst with domain expertise would add value. Second comes the data model. This states how you should structure the database so that the user gets all he wants into one silo.Stage three is when the implementation chaps troop in. They convert the data model into a software package. Finally, this is transferred to users.

With inputs from Krishnan Thiagarajan

eworld@thehindu.co.in

Send this article to Friends by E-Mail
Comment on this article to BLFeedback@thehindu.co.in

Stories in this Section
In the cross-hair

Full of gas? Not this tale

A taste of freedom!

Change your target, friend

Some more way to go on the delivery

Stay online, all the time

Data recovery software

Retain the 'Einsteins' in your office

Cartoon