Companies have spent the past few years gathering their data into one place. Now that they have those “data lakes” under control, they're looking for ways to solve problems on a more real-time basis.
Among the employees in highest demand are what Paul Kent calls “analytically creative problem-solvers.” Almost every company has mounds of data, but it’s these creative problem-solvers who use it to make profit-oriented decisions, like which product to make, what price to charge, or when to send someone a coupon.
Paul Kent thinks about these star employees because, as vice president of big data initiatives in the R&D group at SAS Institute, his job is to look down the road and anticipate which kinds of creative problems to solve next with all that data.
Kent sees big opportunities in two broad areas: speed and the cloud.
Companies have spent the past few years gathering their data into one place. Now that they have those “data lakes” under control, Kent says, “they’re starting to look around and say, ‘Which problems can we solve in more of a real-time basis?’”
Here’s an example. In the mobile telecom industry, especially in Europe, subscriber plans let customers top up their data limit each month if they’re close to going over. So if a loyal customer is getting close to his or her data limit, should the provider offer a free top-up, reach out with a new plan with more data, or just let things be?
“It’s an example of a more real-time differentiated campaign,” Kent says. SAS’s data analytics software helps problem-solvers tune that decision, balancing data elements that include how far into the month a customer is, the customer’s past data usage, and how valuable the customer is to the provider.
The Cloud Could Give Data a ‘Split Personality’
The agility needed to make these near real-time data decisions leads to the second challenge Kent sees for analytical problem-solvers: what to do with data stored in the cloud.
Initially, companies pulled all of their in-house data together into those data lakes so that it’s available for analytics and business intelligence from a single system. “Now, they’re starting to do things that are more born in the cloud,” Kent explains. So, instead of retail customers just having a plastic keychain card they swipe for a loyalty program, they have a mobile app. That app generates more data, “and by the way, that data starts in the cloud, why not leave it in the cloud?” Kent says.
Kent thinks data analysts will start having problems when their data lake itself has a “split personality,” with part of the data in on-premises database software and part of it with some cloud service provider. “How do you handle it when your train of thought takes you to where some of the answer is in your on-premises data lake but some of it is in the cloud part of your data lake?” Kent asks.
Those companies need a hybrid view of their data, he says, offering the example of a company running its SAS data analytics applications on Oracle Database, which some of the world’s leading banks, retailers, and telecoms do.
Oracle lets companies create identical database and infrastructure environments on premises and in the cloud, since Oracle's Database Exadata Cloud Service is identical to the Oracle Database Exadata Cloud Machine that a company can run on premises. And with Oracle’s Cloud at Customer offering, a company can have an Exadata machine on premises, but pay for it monthly based on usage, while having it maintained by Oracle exactly as it would for a cloud service.
SAS is helping companies with these choices, in close collaboration with Oracle development teams. SAS customers can run its analytics software on Oracle Database Exadata Cloud Service, and SAS itself is running its software on Oracle Cloud—for services including database, infrastructure and backup—on behalf of some customers, as a managed service.
Kent thinks having identical environments in the cloud and on premises will help companies looking to move their data analytics workloads, bit by bit, to the cloud.
“People move to the cloud over time,” he says. “As long as it’s easy, people can pick up one thing, then another, then another. If the environment is different, you probably have to be more disruptive as you move. I think you can be more incremental and less disruptive if it’s the same infrastructure.”
The tools and tactics are changing, but the goal is to give a company’s “analytically creative problem-solvers” the tools to learn what drives customer behavior. Says Kent: “From a business standpoint, the goal is to give that inquisitive person a place they can hunt and fish for what is it that influences a decision.”
Chris Murphy is director of cloud content at Oracle. This post originally appeared on OracleVoice on Forbes.
Guest blogs such as this one are published monthly and are part of MSPmentor's annual platinum sponsorship.