Search the Internet          Tips  Sites: Finance, Investment

Data mining: New business opportunity for IT in India

M D Riti in Bangalore

Finally, it's out in the open. As the slowdown hits the IT industry globally, India has dropped its pretensions of becoming the place that spawned the next Microsoft or IBM. Instead, it is trying hard to sell itself as a global back office.

As IT companies in the US find it increasingly hard to sustain their large overheads, infrastructure and staff, the Indian software industry is offering itself as a great new place that can run just about anybody's back office effortlessly.

This is probably the right moment to try this approach as McKinsey predicts that the global market for IT-enabled services will touch $140 billion by 2008. India is desperately trying for at least a modest $17 billion of this huge market.

Interestingly, the largest chunk of this business, will not be the customer interaction services (read call centres) that have been the high profile end of this segment so far. It is data search, integration and management that are expected to account for 44 per cent of this segment. The next, of course, will be customer interaction services.

Not surprising, then, that the CRM (customer relationship management) companies in India, who were so far quite content to set up huge call centres and wait for the money to pour in, have now started separate wings for data mining.

"This dimension of our work now accounts for 15 per cent of our business," says Aashu Calappa, general manager, Customerasset.com.

The concept of data mining as a service is still relatively new to India.

Data mining is the search for relationships and global patterns that exist in large databases, but are 'hidden' among the vast amounts of data, such as a relationship between patient data and their medical diagnosis.

These relationships, if understood properly, impart valuable knowledge about the database and the objects in it. If the database is accurate, then it acts as a mirror for the real world that the database represents.

One of the main problems for data mining is that the number of possible relationships is very large. This makes it difficult to detect which are the correct relationships by validating each of them.

This can finally be done best by intelligent search strategies, as taken from the area of machine learning. Another important problem is that information in data objects is often corrupted or missing. Hence, statistical techniques should be applied to estimate the reliability of the discovered relationships

Data mining also goes by the name of Knowledge Discovery in Databases. Technically defined as the nontrivial extraction of implicit, previously unknown, and potentially useful information from data. It uses machine learning, statistical and visualisation techniques to discover and present knowledge in a form that is easily understood.

The chief goal of this exercise is to extract useful information from vast amounts of accumulated data, which may otherwise be somewhat vague at this moment.

It then tries to maximise the amount of useful information extracted. At the same time, of course, it would also try to minimise the loss of productivity resulting from wasted time and resources, prohibitive access or transmission charges, and retrieval of meaningless, incorrect, or corrupted data.

Just how do you represent/store all these large information data sets efficiently? And how do you either extract or transmit material across a distributed multi-user network? Finally, how exactly do you use fast search strategies and fast pattern algorithms to interpret the information you have collected? These are some of the issues confronting firms getting into the data mining business, which is relatively new to India.

However, most Indian companies getting into this line are already into better established and lucrative customer interaction areas like call centres.

Data mining and call centres actually support each other in many ways. The information gleaned about customer buying patterns or behaviour, about common complaints or behaviour patterns, obtained from call centres, help the customer companies handle people better.

Customerasset.com, for example, cites a concrete example of exactly how the data compiled by complaint calls and enquiry calls made to a call centre operating at his company's premises led a well-known online music store to reallocate its advertising spread. The company also changed its purchasing process from paying online through credit cards, to paying on delivery of the purchased goods.

The starting salary in this area of IT-enabled services would be about Rs 6000 to Rs 8000, which still works out very cheap for an American company ($120 to $200 a month!) looking to outsource this rather laborious and tedious job.

The engineering skill required here is little more than the average call centre worker needs to have. All that is required of him is the ability to use search strategies and statistical analysis tools effectively. Still, there would be a cost reduction of almost fifty per cent.

Running a call centre or a whole CRM company is not an easy task, though. For the few success stories, there are many who tried their luck in this business and went bust in a short time.

That's because these places require a lot of investment in terms of infrastructure for operation and training. The situation is entirely different in the US, where housewives even do telemarketing as a part-time job from their own homes.

Back in India, even though there maybe a large English-speaking workforce, these people have to be trained in American accents, usage and etiquette.

Meanwhile, CRM companies continue to take great pains to have offices that look good and are comfortable, even plush, to help fight attrition in employees and also reassure Western clients that they are reliable and well-maintained places.

"A popular misconception among Westerners is that the Indian companies are able to offer very low competitive rates because they don't treat their employees well and that most of the Indian companies, especially call centres are like slave shops," points out Calappa.

"When our prospective clients visit us at our office, they see the bright and lively interiors and the misconception dies instantly. They see and understand for themselves that the working environment here is as good as that in the West."

Sanjay Rao, who has just started working in data mining for a CRM company in Bangalore sums it up best when he says: "It maybe boring work. But if you can make good money without an engineering degree, and in these days of IT slowdown too, why knock it?"

Money
Business News

Tell us what you think of this report