Companies need to make use of collected data

Your telecom service provider knows how many times you've called a number and exactly how long the calls lasted. Your bank keeps tabs on how many times you call the service centre and what your enquiry was. If there is one thing companies have become good at these days, it is collecting data. But it is all this data that companies are sitting on that is emerging as the biggest problem, says Tom Davenport, co-author of Analytics At Work. "Companies have access to extremely sophisticated data, but they are at a loss as to how to make sense of it," he says. "Companies accumulate data and keep it in databases stored underground, safe from nuclear attack — but they aren't using it." However, there is hope. Davenport, who is the President's distinguished professor of information technology and management at Babson College, Massachusetts says that the world is in the middle of an analytics revolution as people realise that raw data can provide them an opportunity to manage more effectively. His earlier book, Competing On Analytics, was about how to build your strategy around analytical capability. But that was rather niche, and over time, Davenport realised that while companies may not want to base their entire strategies on analytics, there was an increasing number who wanted to take decisions based on analytics. "Even now, about 40% of business decisions are based on intuition and gut feel, but there are companies like Google where no decision will be made without research and data analysis," he says. Davenport teamed up once more with Jeanne Harris to provide a 'how-to' framework to build analytical capability. Called the DELTA framework, it explains how to build an analytical framework focusing on data, enterprise, leadership, targets and analysts. While most people focus on obtaining clean data without duplication, Davenport's emphasis is on finding unique data and identifying aspects that could give the company a perspective on business or customers. Further, he says that it is important to have an enterprise perspective on analytics-based decisions rather than focusing on silos within the organisation. A co-ordinated approach would bring in greater focus and consistency, as well as access to various data sources. If the project is being undertaken at the enterprise level, it clearly needs to be supervised by somebody at the top, but it is equally important to empower people at all levels. When undertaking such a project, it is important to decide what aspects of the business are best suited to analytics-based decisions and are likely to be successful. "While every aspect of the business can benefit from becoming more analytical, you cannot apply analytics to everything at once and it is important to set your targets accordingly," says Davenport. The general rule is that if you want to make money, focus on pricing related issues. And finally, there is no avoiding the investment in smart people. "Computers can analyse data, but you need smart people at different levels with different capabilities to make sense of that and tell you how to apply those findings," he says. Even with this structure in place, certain challenges remain. "We are putting sensors on just about every aspect of our environment and each of these generates a lot of data. Analysing this data will be a big issue for organisations. For the longest time, we have been working with structured data, but now, the challenge is making sense of unstructured data," says Davenport. Today, unstructured data accounts for almost 80% of data conversations and this could range from the content of the calls made by the customer to a service provider's helpline to the warranties of white goods and automobiles. Another difficult area is navigating the social media and social networking arena. While these platforms can provide excellent insights as to what customers really want and think of your product, it isn't always easy to understand what is being said. "Most require a fair bit of linguistic analysis to understand what the comments actually mean, whether they are positive or negative. The current level of sophistication is such that beyond basic trends and identifying whether the comment is positive or negative, we cannot glean too much information from social media," he says. However, companies are developing tools for social media analysis and going forward, this is expected to get more sophisticated. Another area where analytics is increasingly gaining importance within the service business is how to cross sell and up sell products to members of the same network who are likely to fit into a similar profile. This is already visible in certain areas like insurance and telecom in India. Long term, there is a shift away from gut-feel based decision making towards an analytical approach, but Davenport says there have been setbacks like Malcolm Gladwell's bestseller Blink! which promotes going by gut feel. "Clearly something went wrong with our decision making abilities, or else we wouldn't have seen a financial crisis of the proportion we did," he says. Well, the global economy's recent track record certainly suggests there is room for improvement.