January 29, 2010 - DNA

Thomas Freidman’s The World is Flat made the case for globalization, rewriting the rules of competition by focusing on Silicon Valley entrepreneurs such as PV Kannan, co-founder and CEO, 24/7 Customer, who set up the first BPO outsourcing operations in India.

In 1995, Kannan founded his first company Business Evolution, bought over five years later by Kana Software for $140 million, after which he started 24/7 Customer, an IT-enabled BPO that provides customer lifecycle management services. “My programming background was on building customer service systems. I thought, what is the impact of the Internet on sales? That’s when I hit on the idea that customer service could be delivered through e-mail and chat, and chat would be a key tool for converting browsers into buyers,” says Kannan.

In an interview with DNA, the California-based entrepreneur shed light on how his firm helps companies deal with customers in the age of Twitter. According to him, companies can use predictive sales software based on advanced mathematical models with online chat to convert people trawling through websites into serious buyers. Excerpts:

Companies are sifting through gobs of digital data that customers feed networks while working, chatting and shopping online to build businesses. How does 24/7 Customer use data and mathematical models to help customers increase sales?

On any given day, big companies receive millions of hits on their websites and the conversions are still fairly low. Visitors come for different purposes, to buy products, or get answers about a purchased product. Can we predict the intention of a visitor to a website? That is the layer that we are peeling with data and mathematical models.

We see a message in a click. If a customer spends two minutes looking at a luxury watch and lives in a rich neighbourhood or zip code, it signifies something. 24/7 Customer is also working on predictive customer service, doing data modelling on customer behaviour, so firms know the issues a customer is likely to have, say, within the first 72 hours of purchasing a product.

One of our business units, 24/7 Innovation Labs, has created models that enable real-time prediction of customer behaviour. It used statisticians to develop advanced mathematical models to predict how visitors to the Adobe Systems website would shop.

They mined data on website traffic, product average order value and online behaviour, which helped to create models that predict who will need assistance, when they will need assistance and then trigger an intervention with a chat for those customers that need help. It offers a way of converting browsers into buyers. Adobe has been using SalesNextT for a year with great results.

We have 14 customers, who use what we call our modelling software; some of those clients also use our online and call centre services.

Are most of your clients American?

Yes, but we do have British, Canadian and Australian finance and insurance companies. The US and UK are our biggest markets. Indian firms have shown interest in our software piece.

Where is the software modelling being done?

In Bangalore. But we do have a team of PhDs in Silicon Valley, who are involved in high-level problem recognition. The design and solution team here comes up with the overall approach, then analysts and statisticians in India develop mathematical models.

With the popularity of Twitter many businesses are focused on how to identify, leverage and respond to customer opinions expressed on social networking sites. It’s against this background that we have launched 24/7tweetview. The product helps companies know what is being said about them online.

Does your firm rely heavily on a business model where it can take advantage of labour arbitrage? We have grown in the classic offshore model. We have close to 8,000 people around the world. Roughly half our team is in India, the other half in Philippines and Latin America. But four years ago, we created Innovation Labs, where a team of software engineers, mathematicians and statisticians together endeavour to be proactive and predictive. In our estimate, about 75% of consumer behaviour is predictable, that is where we come into the picture with our predictive software.

Our customers typically are large firms with 3-5 million customers, which mean they generate heavy traffic. They also get bombarded with phone calls; they sometimes earn revenue from these calls, but some of these calls are just cost. We help them save costs. So we are an IT-based BPO company that relies on the IT to generate profits, not so much on the labour arbitrage.

Are you planning to expand in India?

We are growing 25 to 30% as a company. So yes, we are hiring. In 2009-2010, the growth will be across our centres in India, the Philippines and Latin America. From the couple of thousand people to be added, roughly 750 people will be hired in India and the rest in the Philippines and Latin America