How to Turn a Big Data blunder into Fire to Fuel Future Success?
There’s a saying that goes something like, “All you have to do in order to succeed is to try one more time than you fail.” In other words, like Mom always said, “If at first you don’t succeed, try try again.” As adults, we often forget these simple words of advice, especially when large sums of money are involved. The key to big data success isn’t investing in all of the latest and greatest technologies and hiring the world’s greatest genius data scientist (though those things help), but to learn to turn failures (and there will be failures) into learning tools to fuel future success.
If you’ve failed with big data, you’re in great company. Google failed miserably in predicting flu outbreaks, while Fannie Mae and Freddie Mac analysts blew mortgage default data big time and caused the housing market crash of 2008. Less famously (though inarguably, far dumber), an Amazon T-shirt seller bumbled big data badly enough to auto-generate T-shirts reading, “Keep Calm and Rape a Lot.”
Clearly, there are lessons to be learned. How can you turn a big data blunder into fire to fuel future success?
The Crash and Burn: Focusing on the Technology Instead of the Organization’s Needs
All the right technology gets you nowhere without the direction and purpose to drive initiatives.
Granted, the technology here is razzle dazzle. But if you fail to uncover issues that need to be resolved, or neglect to use pertinent data to forward the organization’s interests, there is no real return on the big data investment. It’s not really beneficial to discover what your most popular product is or how many people use your printed coupons if this data doesn’t serve a real purpose, either in production or product development or marketing.
The Fuel for Future Success: Determine the Needs and Goals of Big Data, Then Use Technologies to Solve the Problems
Start with the end in mind. Big data will give you directions to get where you need to go.
Look at the business’ needs first, then determine what data can offer insight into better practices. Don’t even crack open the databases until you determine what the data needs to do for you. Is there too much waste in the production process? Too few Midwestern homeowners buying your product? Not enough Millennials using your mobile coupons? Find out what the problems are — then analyze your data to uncover the answers.
The Crash and Burn: Centralizing All of the Data Just Because
Big data is really, really big. A common mistake is for organizations to pool all of their data resources together, making it too unwieldy and cumbersome for any single department to make use of it. Complete centralization of data leads to too much information for anyone in the organization to use, and not enough information for any one department to use.
The Fuel for Future Success: Make Relevant Data Available Where Needed
A much better way to approach big data is to structure data warehousing so that each department has their own sets of data and purpose for using it. Then, each department is able to yield meaningful results for their purposes. Naturally, the marketing department will uncover useful information to pass on to R&D, and R&D will discover purposeful information to pass on to production managers. But everyone has what they need, and only what they need for timely, relevant analysis.
The Crash and Burn: Running Up Costs With In-House Data Warehousing Infrastructure
A huge myth of big data is that it’s too expensive to undertake. It actually isn’t, so long as you make use of the low-cost resources available. Big data only becomes too expensive when organizations make unnecessary investments in massive in-house data warehouses. Often, this is done even before the organization has determined exactly what the data is supposed to return to them for the investment.
The Fuel for Future Success: Use the Cloud for Greater Data Availability, Cheaper Operations, and Easier Access
By taking advantage of cloud services, organizations can house, manage, and analyze massive data sets without all of the infrastructure investments. This means the data can begin returning value on investments quickly. Additionally, cloud storage makes the data available to remote workers, which is critical in today’s mobile world. Failing to provide mobile access to data is one of the significant hindrances in organizations’ ability to achieve return on big data investments.
Instead of scrapping a big data project that seemingly failed, simply reanalyze how, where, and why you’re using the data, and figure out a way to fuel success in the future using the lessons you learned in the past. After all, isn’t that what big data is really about?