Big Data Goes to Work
6. Big Data Goes to Work
The competition between big data and traditional enterprise data is over: they both win
Read more about Big Data Goes to Work
Senior Vice President and CIO
Like any other large business, our company has amassed enormous amounts of data over the years – most of which was generated in a fairly structured environment. But today we’re gathering new kinds of data from a wider range of sources, including in-house customer information, meteorological data, supplier data, census information, other government sources and even the “digital exhaust” from social media. It’s virtually all focused on two business imperatives: growth and performance.
Demands for better information and more insights come from the business: finance wants profitability reports; the dairy business asks for improved forecasts; the seed division needs better information for the salesforce. You get the picture. We’re even analyzing satellite imaging data about individual farms and fields, and combining it with geological data to help farmers make seed choices that can improve their yield. But the picture isn’t entirely rosy. Bottlenecks are virtually everywhere. In some cases, a business leader may not realize what kind of questions he could be asking. Technology hurdles are real, especially in a diverse landscape like ours. And the talent needed to deliver big data and advanced analytics is in short supply. From technology platform experience to statisticians to business analysts, it’s difficult to hire and develop our people fast enough to stay ahead of our information needs.
However, taking advantage of big data isn’t primarily a technology issue; it’s more of a mindset issue. Big data has given us the opportunity – and the need – to start thinking in new ways.
That said, we do have an expanding suite of tools for business intelligence and analytics. One of my top priorities is to standardize around a common toolset that can meet our dashboard and reporting needs – and also handle the growing amount of unstructured data flowing through our enterprise.
How do you quantify the value of being able to do this? It’s easy when you’re focused on the right outcomes. For example, by examining seasonal buying patterns among our largest customers, we were able to take three days out of butter inventory. It required a massive amount of analysis – and it was worth every penny. In some cases, we’re even monetizing our analytics capabilities by selling information to large customers and distributors.
Bottom line? If we can use big data to help a farmer go from 160 to 180 bushels per acre, we’re doing something really valuable. When our customers can deliver more yield, we can all win.
Where do you start?
Big data is not an all-or-nothing affair. Companies today can consider a multitude of new data sources with volumes, varieties and complexities that would have been unfathomable just a few years ago. It is easy to get lost in the possibilities and become overwhelmed by the enormity of the implications. To stay focused, find practical entry points to big data that are digestible in scope, and that can enrich your current analytics journey. Along the way, keep the following in mind.
- Crunchy questions. Start with highly specific questions and total clarity around the business problems driving the journey. Those – and only those – data sources deemed important and linked to the business objective should receive care and feeding. That way, big data can go to work generating the desired results that inspire broader potential initiatives to pursue.
- A 15-degree view of the customer. Attempting to inventory, cleanse and manage the breadth and scale of big data is impossible. The answer is not a single, unified, canonical data store across virtually all sources and types. Instead of trying to take broad view, determine the two or three discrete items that can shed light on the targeted crunchy questions. Fiercely protect scope around these areas, leaving the other 345 degrees of visibility untouched until needed.
- Nimble governance. One of the challenges of big data is the need to be agile. The structure of incoming data is often not known in advance, or it can change over time. New data sources may be added. Different business outcomes require data at different levels of accuracy, granularity and availability. Tradeoffs must be made in cleansing and de-duplicating data, models and analyses – quality versus responsiveness. Standards and stewardship can play a role, as can transparency of the source and visibility into confidence levels about data accuracy and quality. The artistry comes in fine-tuning the rigor of governance with the need for agility.
- Co-existence. Avoid religious arguments between conventional analytics teams and big data advocates. Most leading organizations can use both. Some education may be needed to establish an understanding of the most effective applications for each domain. Just like the debate between OLTP and OLAP, the answer will likely be mutually beneficial co-existence, at least until hybrid offerings make the argument moot.
- Avoid vendor churn. The vendor landscape around big data players is huge, with the potential for consolidation possibly years away. Credible, compelling offerings may exist at each tier – from the biggest players to start-ups. There is no specific answer, though there are important differences in technology, approach and delivery models. Balance cost, needs, the ability to leverage existing vendor relationships and contracts and the appetite for experimenting with open-source or cloud-based solutions. The big data world is rapidly changing. Prioritize rapid experimentation over drawn-out vendor comparisons, especially when getting started. Wait to make more strategic, binding vendor decisions until the goals and significance of big data are understood.
The explosion of big data was likely inevitable. What once was the stuff of science fiction has become an everyday occurrence. Exabytes. Zettabytes. Yottabytes. Sensors, asset intelligence, mobile devices and constant streams of unstructured communications have created digital exhaust that can capture who we are and how we live, work and play. Powerful new insights can be gleaned, with the most value realized by those who learn to detect signals from the noise, derive meaning from the signals and turn meaning into action. Focus on specific crunchy questions tied to well-defined business problems and attributable results. Organizations that put big data to work may pursue a huge competitive edge in 2012 – and beyond.
As used in this document, “Deloitte” means Deloitte LLP and its subsidiaries. Please see www.deloitte.com/us/about for a detailed description of the legal structure of Deloitte LLP and its subsidiaries. Certain services may not be available to attest clients under the rules and regulations of public accounting.