Big Data – Unified analytics: A new mastery over the data wave
By David Zakkam and Ankesh Aggrawal (l-r)
Data is exploding. And it is not necessarily a good thing. Organizations feel like they are “drinking water from the fire hose.” There are difficulties, not only in managing data but also in trying to generate benefits from the data. Organizations are now starting to realize that today’s world is not about squeezing the most juice out of data but about trying to separate the tiny signal from all the noise. The need is to work backward from the outcomes that need to be achieved to the behaviors that lead to those outcomes and to the insights that would lead to changed behaviors. The focus should be more on asking the right questions rather than seeking the best possible answers.
The best insights are generated not from answering one question but from answering a set of connected questions. Answering these connected questions requires using a connected set of data sources. It requires us to make a paradigm shift – from looking at each faction of data as an individual unit to viewing data from multiple sources as a single symbiotic entity. Unified analytics is an upcoming field that throws light, not just on the nuances of the data, but also on the interaction between data and insights. It will enable the access and analysis of data from multiple sources on a single interface. To help you get started in this field, here are three things to keep in mind:
1. Integration of unstructured and structured data.
Integrating structured and unstructured data is one of the most important tasks required to run unified analytics. Structured data from transactional systems enable understanding of what a customer is doing. Unstructured data from sources like blogs, videos, discussion forums and call center discussions allow businesses to understand what a customer is thinking and feeling. Looking at these data sources in an integrated manner creates the all-important link between “what the customer is thinking” and “what the customer is doing.” Any analysis done using integrated data would have far higher quality of insights than doing analysis on either of these sources alone.
Although it is not easy to integrate unstructured data into traditional environments, organizations don’t have much of a choice. As per Gartner and IDC, unstructured data will comprise more than 80 percent of the 7.9 zetta-bytes of data that will be created in 2015. Therefore, organizations are compelled to make the leap towards integration of unstructured data.
2. Looking beyond traditional sources within the firewall.
Organizations have traditionally looked at utilizing data from within their own firewall. However, there is a wealth of information outside the firewall. Imagine a clothes retailer trying to figure out whether to put the winter clothing line on sale or not. Today, the retailer can purchase weather information from third parties and estimate fairly accurately when it would not be cold enough for winter-wear. That information would be useful in figuring out the sales expected, which would help out in estimating stock required. This “inter-firewall” analysis could be a competitive advantage, but the window of opportunity is quite short.
There is also plenty of data outside the firewall that is openly accessible. Macroeconomic data for most of the world is available from the United Nations websites, the U.S. census information is available for download online, and 500 million tweets per day are available for analysis. This “trans-firewall” data can also become very valuable to enterprises if leveraged optimally and in real time.
In the near future, the number and diversity of data sources is going to explode. The Internet of Things phenomenon will generate enormous amount of very interesting data. The human body may become the biggest data source out there. There may be many other data sources that we may not even be able to imagine today but may become a reality very soon. Organizations need to quickly start looking beyond the firewall and explore the various data sources outside the firewall.
3. Customer focused data strategy.
Steve Jobs said, “You’ve got to start with customer experience and work back toward the technology – not the other way around.” A company’s data strategy should also start with the customer. A customer today has a number of touch points with the company, and the touch points continue to increase with the widespread use of technology. Customer experience is at play throughout the customer journey: while evaluating a product, during the purchase process and during consumption. Creating a 360-degree view of the customer therefore is the first step in the journey toward providing a better customer experience.
To face the challenges of today, companies need to quickly start looking at the benefits of exploring and utilizing multiple data sources to achieve business outcomes. It is definitely a challenge but moving ahead is the only option. Companies need to look inside as well as outside the firewall at the data sources that are freely available or procured externally. They need to look at exploring and working with large amounts of unstructured data that is difficult to store, process and analyze. They also need to keep the customer at the center of their data strategy so that the customer experience is the best possible.
David Zakkam is an apprentice leader at Mu Sigma and has 12 years of experience working in the analytics industry across areas such as CRM, social media analytics, rapid impact analytics, search engine monetization and big data technologies. He is currently involved in studying the interconnectedness of business problems and helping organizations benefit from complexity. Zakkam has an MBA from the Indian Institute of Management, Calcutta, and an engineering degree from the Indian Institute of Technology, Delhi.
Ankesh Aggrawal is a senior business analyst at Mu Sigma. He has worked on regression techniques, market-mix, machine learning techniques, text analytics and social media analytics. He has a bachelor’s degree in engineering from Punjab Engineering College, Chandigarh, India.