Thursday, November 15, 2012



Bharti Airtel handles around 8 billion calls daily generating humongous amount of data. Based on this it has divided its customers into nearly 100,000 segments to whom it offers customised products based on their usage pattern. Right now it is working to provide at least three services to each customer, to improve stickiness, working on the insight that customers for multiple services tend to stay on longer than single service users. Further, due to call drops it was facing the problem of churn. At nearly 40% it was a huge number. When it analysed the data it found that a customer was most likely to switch to another service provider if six or more calls dropped in a day. So a software was developed that offers customers free SMSes after the sixth call drop.

British Airways (BA) cabin crew and ground staff are now provided iPads to store, transmit, and use real time info about who is on board, their needs and expectations, preferences, and even any issue that they might have had at the airport. It realised that the premium class flyers feel delighted if they are addressed by their name. So the crew now does exactly that thanks to transmission of real time data through the handheld iPad. While other airlines can at best have current data on their passengers, BA would have the complete service history of them which it uses to draw patterns. Welcome to the world of analytics and its latest incarnation Big Data.

Analytics involves sourcing data from internal and external channels (media, blogs, image, and video sites, even sensors, and CCTVs), organise this data, generate insights, make predictions and finally produce recommendations for action. Big Data goes even beyond. It refers to a collection of information, too large and complex to be processed using traditional software tools. This processed information in turn helps the company identify untapped revenue potential, generate insights to cut costs and boost profit. For example, McDonald’s is able to understand where to locate its stores, how to layout these stores, which items to keep in each store and which items to bunch together to generate bundled sales. Shopper’s Stop got the insight that Gujaratis’ purchases are closely linked to upswings on the bourses since they would spend their gains from the stock market on purchases in the retail. Shopper’s Stop realised 25% incremental sales from them. During Eid it searched through its database of 2.6 million First Citizen loyalty card holder and singled out Muslim buyers. Then through targeted promotional programmes it earned Rs.1 crore additional revenue. Jet Airways is able to accurately calculate, track, and report aircraft emissions, so as to optimise its fuel usage for each flight. Little wonder then that an IDC Report tells us that organisations which apply analytics to data for competitive advantage are 2.2 times more likely to substantially outperform their industry peers. Companies adept at analytics enjoy 1.6 times more revenue growth, 2 times more profit growth, and 2.5 times more stock appreciation than their peers.

Marketing Matrices
According to India Retail Report (2013), private consumption in India increased from Rs.44,11,115 crore (2011) to Rs.51,26,131 crore (2012). Between 2010-2015 the estimated CAGR is likely to be 18.8% for retail sales. Consequently the share of modern retail will move up from 6.6% to 10.2%. While the total retail is likely to grow at 16% annually, modern retail will hit a figure of nearly 27%, a big jump from Rs.2,23,572 crore to Rs.4,87,423 crore during 2012- 2015. But then this will invite a lot of competition too, since the entry is easy. Who will then breast the tape first on the finishing line and who will be disqualified in the preheat rounds itself? Well, inter alia, that would depend on how customised your programme could be to suit individual customer needs, and how tightly you run your ship. Increasingly, the segment size is likely to be a single customer at a time. You need to process zettabytes of data (1 Zettabyte = 1,099,511,627,776 Gigabytes). Hence, the marketer will need the support of several matrices (Table 1).

Simply defined a metric is a measuring system that quantifies a trend, dynamic, or characteristic. Marketers need to understand their addressable markets quantitatively as well as qualitatively. They must measure new opportunities and the investments needed to realise them. They must quantify the value of products, customers, and distribution channels all under various pricing and promotional scenarios (Table 2).

Put together these matrices are designed to measure how well the firm is doing with its customers as a whole and wherein lies the scope for enhancing profitability of operations.

Share Matrices
A word of caution, however. Reams of data (or zettabytes of it, if you so wish to call them) are of no consequence unless they generate real insights. In Philippines diesel and petrol have similar price differential as in India. An automobile company was about to make the blunder of targeting the Filipino market with India made diesel cars. However, before the disaster could have struck it, it found out that the natives there were not too fond of diesel engines; 86% prefer to buy petrol driven cars. Reason: Since distances are not long, a diesel engine does not offer overall economy. Similarly, in the early to mid nineties, many Indian companies betted on CRM software. But most such initiatives flopped because these were not implemented in sync with the company’s processes and decision making norms. Tesco, which will soon retail in India, has Clubcard loyalty programme under which it offers highly customised coupons; hardly any two mailings have the same coupon. An Indian apparel major has divided its customers into over 200 micro clusters, each based on distinct behaviour. The retailer configures the segment properties and engagement content and then makes the right offer to the right product at the right time.

Remember, as always, technology is a mere enabler; it is the human ingenuity that counts.