More data than ever is flowing into enterprises from websites, smartphones, social media, wearables, video, cars, contact centers, ERP systems and the expanding IoT ecosystem. Yet being data rich and information poor is not a good thing. In today’s increasingly connected and competitive economy, enterprises need immediate and real-time access to data to make smart business decisions. They also need to know that their data is secure, accessible, trackable and useful. Data—and how enterprises use it—will be an even bigger differentiator in 2017 than it was last year.
Here are three major areas where we predict enterprises will see big advances in terms of data:
1. Security and cloud become synonymous
The consensus used to be that cloud computing was less secure than on-premises computing. That is no longer the case and enterprises will make big moves to the cloud with their data in 2017. Gartner estimates that more than 50 percent of global companies already store customer sensitive data in the cloud. That will only grow. No doubt, security will remain a concern. Hackers and cyber thieves will only get better at what they do. To combat that, DevOps teams need to incorporate security best practices and proven technologies into their everyday practices. It can’t be an afterthought. When it is an afterthought, bad things can happen. Instead, enterprises will look at all the layers of IT infrastructure with security in mind—whether that’s on-premises within the actual “four walls” of their own physical data centers or in the public cloud within the “virtual four walls” that they create. Public cloud vendors now provide all tools to enable data security at multiple layers. Yet it will be up to the enterprise to make use of those tools. Mature enterprise software should provide all of the hooks to allow companies to leverage security at every layer of a complex IT infrastructure, not just assuming that an “around-the-perimeter” strategy will be sufficient. This is why data level security is of particular importance. It’s like securing a house. You can have security cameras on doors and motion sensors on windows. Yet if intruders still get into the house, you’d prefer that they find a 300-pound safe bolted to the floor instead of a full jewelry box. That is how enterprises must think about securing their data.
2. Governance wins popularity contest
Historically, data governance has not been a hot topic. Instead, it’s often been viewed as something enterprises have to do for compliance or regulatory reasons. Yet as data drives more business decisions, enterprises will see that good data governance is crucial to driving business value. They’ll implement projects faster, with less risk and less ongoing cost. You can’t get consistent, reliable and repeatable data without data governance. Engineers would never build a bridge without blueprints, and analytics teams need data governance to guide and structure activities. To truly trust data and the actions and insights you get from it, you have to know where that data comes from and when and how it was collected. That is more of a challenge now than ever. In addition to structured data, which often includes things like demographic data or personally identifiable information, enterprises are now receiving tons of unstructured data, such as social media comments or notes from contact center agents. Data governance is crucial because it ensures that everyone is speaking the same language when it comes to the meaning of the data.
Enterprises deal with governance at multiple layers, from a conceptual layer (e.g., what are the business entities, what are the relevant regulations, etc.?), to a process layer (e.g., how do we manage the governance program itself?), and into the application portfolio affected by the governance processes. Below all of that is the data layer itself, and more often than not, governance programs focus a lot of energy on the conceptual and process side of governance, with perhaps not as much focus on the concrete things that actually happen at the data layer, sometimes resulting in a disconnect between the governance goals and the actual execution. Organizations however that can execute good governance at the data level will be far ahead. One of the ways that they can do that is to provide mechanisms that allow for data to be safely and securely enriched in place, so that data governance rules may be applied with as little friction as possible. Take for instance, personally identifiable information (PII). All enterprises need to secure and track that. At the data layer, that means you’ve got to account for all the myriad ways that PII might be represented in the data. It might be a social security number that fits a certain pattern in a certain field but it also might be buried inside of other data. Good data governance means having a data layer that embraces the multiple representations of PII that are possible and then being able to allow those multiple representations of data to be enriched in place without having to try to conform to a non-existent “perfect” model.
3. Databases get smarter
IT departments are under tremendous pressure to deal with a perfect storm of data requirements. All these pressures point to one outcome: Databases will get smarter to make operations faster, provide better answers and even provide unexpected insights. Enterprises today typically have hundreds, even thousands, of data silos. Each contains important information, but is unable to provide a complete picture. For years, we’ve been integrating data from different silos—but at a cost. The relational database technology that’s dominated the market for three decades requires enterprises, when integrating data, to do so from the lowest common denominators. They’ve been forced to “leave data on the floor” when trying to harmonize disparate models and bury the context around the transformation of that data inside complex code. As a result, they end up with less of their data and lack of context. With next-generation database technology, Enterprise NoSQL, the provenance and lineage of data sits next to the data, not in code somewhere else, so enterprises have ready access to the context of the data and how it relates to data from different silos. That’s where the smarts of the database come in, by harmonizing seemingly disparate pieces of data and making information contextual in real time. These kinds of smarts would, for instance, enable an enterprise to predict—and then proactively correct—the impact on customers if, for example, a single network switch deep inside the enterprise should falter. By having all of the contextual richness—things like business purpose and customer dependency—all in the data layer and making it available to the all the right people at the right time, enterprises will do great things.