There’s no shortage of clichés for why enterprises should care about data, but talking about the importance of data and making it useful for enterprises are two very different things.
THIS: Data governance checklist for your organization (TechRepublic Premium)
Over the past few years, data democratization has become a rallying cry for enterprises hoping to make better use of their data, and vendors hoping to sell magic solutions towards that goal.
What is data democratization, and how can enterprises tap into this trend to hone and better leverage their employees’ data practices?
What is data democratization?
Data democratization is often wrongly assumed to be all about access; that is, some suggest it’s for ensuring that “every user in the organization, regardless of their technical prowess, can have access to data for timely and more insightful decision-making.”
But access is only part of the equation, and not the most important part, as historical founder, Arpit Choudhury, has correctly called out: “Data democratization is the ongoing process of enabling everybody in an organization, irrespective of their technical know-how, to work with data comfortably, to feel confident talking about it, and as a result, make data-informed decisions and build customer experiences powered by data.”
In other words, it’s not really about making data available to everyone, but rather actively enabling employees to use data well.
Data democratization tips and best practices for your business
Since data democratization is about both access and appropriate training, what steps does your organization need to take in order to increase data democratization and ultimately benefit from it?
Open up data access
Traditionally, enterprises have kept data locked down and accessible to only a chosen few. While this may seem like a reasonable way to ensure security, it’s a terrible way to facilitate data-driven decisions. As Saul Judah, VP analyst at Gartner, has argued, data governance must embrace “speed and agility,” which has rendered “traditional approaches to data governance … obsolete.”
As such, companies increasingly depend upon tooling that opens up access to data for faster, more efficient data-driven projects across the organization.
THIS: How do I become a data steward? (TechRepublic)
To make data more broadly accessible, data governance tools are increasingly incorporating policy and stewardship management capabilities, thus simplifying access for a wider variety of user roles. Data governance tools also often have AI/ML and associated capabilities built in from the start. Finally, and importantly, organizations increasingly rely on data catalogs from vendors like Alation to constantly crawl through an enterprise’s data and make it broadly searchable, Google-style.
While enterprises need to ensure that data professionals like data analysts or data scientists have access to data, true data democratization is recognizing that every role can be improved with heightened access to useful data.
Improve data literacy
As much as we like to trumpet technology as the fix for data democratization, it’s perhaps even more important to focus on improving employees’ ability to read and understand data. True, this is partly a technology issue, but it’s much more than this.
Data literacy is the ability to read, work with, analyze and argue with data. Enterprises should emphasize and improve their employees’ ability to understand and use data — not just access to the tools used to ingest or analyze that data.
As quoted in an MIT Sloan School of Management analysis, Cindi Howson, chief data strategy officer at ThoughtSpot, stressed: “If we’re spending 80% now on technology, 20% on data, flip it — make the technology super easy so that you can spend more time on data.”
THIS: Review the top data literacy training courses (TechRepublic)
Clearly, the basis for any good data literacy plan is a strong focus on data fluency, not just technology. How do you do this?
The first step may be to establish a data skills academy within your organization with executive support. Rather than attempting to instill a general-purpose vision of data’s importance, the program should be tailored to the particular needs and data sources of a given enterprise, its departments, and specific roles and responsibilities.
In a similar fashion, the company should use examples that are cross-functional in nature, making it clear how data can be useful across the enterprise. Although some data analysis skills like statistics or lookups can seem daunting, emphasizing how to use them successfully can make learning about them feel more manageable and less nebulous.
Data literacy training becomes much simpler when enterprises ensure that data is a critical and obvious part of decision-making. The more often that conversations about data can be made pertinent to a particular organization’s needs, the easier it will be for employees to build their knowledge and invest the necessary time into becoming proficient users of company data.
Turn your data warehouse into a data lakehouse
For years, companies turned to data lakes to store all their data in a central place. However, too often, data lakes turn out to be data swamps that are plagued by poor governance and security. Rather than dump disorganized data into a data lake, many enterprises are now completing extensive data cleansing and preparation projects prior to adding their data to data lake environments.
THIS: Hiring kit: Database engineer (TechRepublic Premium)
Coupled with a shift to more of a data lakehouse architecture, this can help enterprises make data more readily useful and available to employees of all backgrounds. A data lakehouse adds a transactional storage layer on top of the data lake; rather than ETL-ing data around, organizations can apply data analytics services like Amazon Redshift or Google BigQuery to data in place.
Embrace the cloud
Although the cloud still represents a relatively small percentage of overall IT spending, it makes up a significant share of net new IT spending. Directionally, it’s where the greatest share of IT spending is heading and, by extension, where an ever-rising share of corporate data is born. This is both good and bad for businesses.
The bad news is that many enterprises have yet to get up to speed with cloud technology. As indicated in Zaloni’s survey data, few enterprises feel they have the “skills needed to manage new cloud technologies”.
In fact, it’s the single-biggest impediment to data governance success, according to these survey respondents. So the cloud is one of the biggest trends — and potential roadblocks — to data governance and democratization today. For enterprises that want to truly democratize data, it will be important to teach employees how to effectively use cloud-based tooling to probe cloud-based data.
Why is enterprise data democratization growing in importance?
Given the speed at which enterprises must operate, it’s no longer sufficient for data analysis and knowledge to belong solely to IT or some specialized group. Yes, data scientists and other experts will be able to complete different and more advanced data analysis projects, but for a company to thrive at scale and speed, organizations must open up access and utility of their data to all employees.
That doesn’t mean someone in the sales department needs access to the data that’s used to make smarter facilities decisions, but it does mean the default should be more and better data fluency across the larger organization.
Read next: Top data quality tools (TechRepublic)
Disclosure: I work for MongoDB but the views expressed here are mine.