5 Steps to Manage Risk in a Big Data Environment
Not much of consequence happens without risk. As more organizations realize the value of Hadoop while they look to adopt big data into their technology portfolio, they also need to consider the inherent potential for negative consequences. Big data has opened up a whole new world of risk, but that’s not stopping — or even slowing — many businesses looking to cash in on the rewards. To balance this process, technology and business leaders should know how to manage the conversations around big data risks as well as rewards.
When viewed through the lens of risk, organizations have different classifications and considerations to own:
1. Data security and administration
Data security and administration are the obvious issues that usually get the first look. But there are many technical layers for appreciating the security of your data, including:
data encrypted at rest and in transit
proper configuration for authentication, provisioning, onboarding, offboarding
high availability and failover
bare metal versus cloud
Who is going to manage this environment? Can you find the talent to stand up, lock down and maintain your big data stack? When new big data initiatives are launched, these questions are the first things that IT and your information security team will want to know. Be ready with the answers, and know why these things are important for securing funding and buy-in.
2. Data governance
How do you manage the ingress and organization of the data? One of the hidden risks of a comprehensive data lake is that data from one source can be combined with data from another to create inadvertent data exposures. The unforeseen downside of bringing in all the data to one place is that existing controls and processes for privacy may be obviated. Data governance is more than that, of course, but be ready to have a data governance strategy, embrace it and partner with your data stewards early in the process.
3. Speed-to-market philosophy
Is there a cost to NOT having the tools in place, like not being able to leverage your data assets? This is a new technology landscape – business analysts have to learn how to hunt for their own data. The onus for coding business rules into viable code has shifted responsibilities from process-heavy IT functions to results-oriented business units. With great power comes great responsibility, but you should trust your people and reward them with your “data first” ethos.
4. Are you just recreating existing processes with different technologies?
This is one of the biggest latent risks because it indicates that the technologies have evolved but your mindset has not. It can be like using a hammer to drive in a screw. You just spent a lot of money to recreate your data warehouse in Hadoop – and that’s not what it’s for. Understanding the differences between a data lake and a data warehouse will be important, and be ready to preach this on a daily basis.
5. What if this is all a fad and the traditional vendors catch up?
There are vendor management implications, for sure. Maybe it would just be easier on procurement if a database just released their own big data stack? Unfortunately, that’s not how this works. Organizations need to accept that big data environments are complements to their existing technology stack, and that the new players are approaching data analytics from a different perspective.
The Big Picture
Organizations need to understand – if not obsess about — the relationships between their big data environment and the inherent risks associated with having or not having one. Innovation will not arrive without risk, and when thoughtfully managed and understood, your organization will be better prepared to move forward.
The rewards and bounty for succeeding with big data are just now being realized. For some organizations, that means better customer service, retention or acquisition. Profits may improve by creating new, sophisticated product recommendations. For other organizations, fraud identification and prevention techniques are reducing overall costs and isolating additional risk points. All kinds of big data risk/benefit scenarios are emerging, and many companies have concluded that they are ready because they took the time to weigh the risks and convey the “whys” throughout their company. Because if you can’t assess the yield from your big data strategy, you aren’t ready to take that first, risky Big step.