Over the past couple of posts, I’ve talked about what pervasive data intelligence is and why you need it, as well as the technological and human factors involved in making it a reality. This week, we’re going to dive into the environment you need to create in your organization to embed data intelligence in each part of it—to make it pervasive.
At most companies, though they might have sophisticated analytics capabilities, the ability to leverage those capabilities to make better decisions and gain a 360-degree view of their business (I know, it’s a cliché, but it’s true) is often severely compromised. Why? Three reasons. One, their information is siloed and their analytics applications are fragmented and often-extemporaneous and need-specific. Two, their analytics capabilities don’t fit a wide variety of user needs and ability levels, and three, those capabilities are not available on an as-needed, when-needed basis.
In an environment like this, it’s nearly impossible to gain value from your analytics investments. To fix the problem and create the right environment for pervasive data intelligence, you must do three things: unify your information and analytics capabilities, democratize them, and commoditize them.
Most mid-to-large-size companies have an analytics strategy. They generally know what they want to achieve, and they may have a plan to get there, but their plans are often derailed by problems that arise and need a quick, specific—and often dirty—solution. One or two of these cases won’t kill you. A company full of them, though, results in information silos and fragmented analytics solutions that are really just problems.
To unify your analytics capabilities, first eliminate the information silos. That’s a tall order, but you can do it. Get a unified, or unified set, of data stores that feed your analytics engines. Clean your data and put governance policies in place to keep it that way.
Next, get an analytics platform—not just a set of quasi-connected applications. I’m not saying that you have to standardize on one analytics engine. Far from it. Instead, what you need is a platform with an orchestration layer that feeds analytics requests to multiple engines across a high-speed data web so that if a data scientist wants to use a SQL engine, a machine learning engine, or a graph engine, she can do that without moving from the platform. This environment gives you analytics capabilities that aren’t bound by organizational silos and that give you unified answers to your questions, i.e., it makes intelligence pervasive.
Another problem many companies face with their analytics capabilities is that those capabilities are only available to a select group of users. Whether that’s because the capabilities are few in number, or relegated to specific departments, it’s a problem.
A pervasive intelligence platform can help you solve it by enabling data scientists and business users alike to avail themselves of its capabilities: it can democratize analytics. For example, data scientists can use favorite languages like SQL and R, as well as favorite workbenches like Jupyter and R Studio, or any other you might want to employ, while business users could use a drag-and-drop tool like KNIME. Whatever applications your analysts and users want would theoretically be available, as long as they can justify its use vis-à-vis organizational goals. What’s more, those tools will be available in the same environment without making the users jump through hoops to find what they want, when they want it.
Speaking of making analytics capabilities available when and where users need them, unfortunately many companies just don’t—or can’t. Again, it’s a problem of scattered capabilities and fragmented information. But more than that, it’s the spectre of the gargantuan task that is managing large-scale analytics capabilities that scares many companies off from investing in widespread analytics.
But what if you didn’t have to manage all that storage and analytics power yourself? You don’t. That’s what the cloud is for. 70% of your competitors already have at least one app in the cloud, and you probably do too. Why not leverage that and build off of it to make analytics a commodity in your organization?
With an intelligent cloud, you can store all the data—both structured and unstructured—and have access to all the analytics capabilities you need—for a defined cost, and you can have that power managed by someone else to free your IT people and business users up to focus on value-added work. With the analytics commoditization that intelligent clouds provide, your people can access the analytics power they need, when they need it, where they need it. That is the very definition of pervasive data intelligence.
Building your pervasive intelligence environment
Let’s be clear. I know that many companies still struggle to implement analytics. Maybe yours is one of them, and you might be thinking, “Well if I can’t get analytics right in the first place, how can I begin to embed it throughout my company?” It’s like phone service in some countries. Those countries, for whatever geographic or economic reasons, couldn’t build comprehensive landline infrastructures, so when cellular technology emerged, they went from next-to-nothing to full-blown cell service, so you’ll often see the seeming incongruity of remote villagers on iPhones.
It’s that way with analytics too. You can begin your analytics journey with a pervasive intelligence platform. In fact, that’s probably the best way to go. Build it right from the beginning, and you won’t have to burn dollars constantly rebuilding as your needs change. That’s a prescription for analytics success. That’s pervasive data intelligence.