While the age of cloud storage has removed the physical barriers between raw data and consumers of information, it hasn’t removed the organizational hurdles. Many believe it’s time for that to change.
Businesses store increasingly unwieldy volumes of information, much of which can be out of date, inaccurate, or difficult for a layman to understand. The traditional response to managing this mess has been to gatekeep information behind the IT department, where it can be retrieved and processed by request before delivery to its end consumer.
In recent years, a growing number of organizations has sought to break that model, putting data immediately in the hands of many instead of the control of few. This process, called “data democratization,” seeks to eliminate bottlenecks and facilitate faster decision-making, but its success hinges upon its implementation.
The Benefits of a Healthy Democracy
For decades, company information wasn’t accessible at your fingertips; instead, it was managed by one or more data teams. Under that model, the consumer identifies a need for data and sends their request to IT. The IT team then plans the scope of data retrieval, processes the raw information into a more readily-digestible report, then delivers the end product. Often, if the initial request for data is broad in scope, the response time can be burdensome, particularly if the data team has received a large volume of requests in a short time period.
The lag in request fulfillment can be a hindrance to the execution of data-driven corporate strategies. Data democratization can expedite this process, giving customers control over what data they retrieve, and also giving them full access to raw information rather than a processed product. By empowering consumers with this level of access, the organization at large becomes more nimble and more competitive, assuming all goes well.
Democratization also solves scalability issues that come with IT’s management of an ever-growing information database. Advocates for democratization also believe that putting information freely in the hands of employees with diverse expertise will result in the organic creation of novel business strategies, creative modeling solutions, and greater data insights.
Drawbacks and Challenges
Data democratization requires careful implementation to avoid or mitigate its pitfalls. For instance, if not handled properly, free and open access to information can threaten data integrity and make maintenance roles ambiguous. Proactive steps will need to be taken to establish who is responsible for preserving or modifying data.
A more human concern—and one that pervades all aspects of our lives—is the reluctance of subject matter experts to relinquish complex information to a general population that lacks the training, experience, or context to properly interpret it. At the inverse, many end users may actually prefer to receive heavily condensed and processed information, versus sifting through raw data.
The internet age has, in many ways, democratized data across the world, and the results of that experiment are still pending. However, in an enterprise context, measures can be taken to better posture information consumers for success.
There are many methods of democratizing data, most of which are not mutually exclusive. The right mix varies from company to company.
- Enact strong information governance: Before data can be democratized, guidelines should be established dictating who is responsible for data curation and upkeep.
- Employee education: At a societal level, education is one of the central pillars holding up healthy democracies. The same is true in the enterprise context, where employees and stakeholders need to understand data access procedures, how the information is to be utilized, and how to use the tools available to them. Just as important, employees should understand the value of these tools in the execution of their regular tasks. For example, Airbnb has created what it calls its “Data University,” which educates its employees on statistics and analysis, problem solving with data, writing SQL, and data visualization. The result empowered nearly half its employees to become weekly regular users of its self-service platform.
- Embed data analysts: Each team can be assigned its own data analyst, dedicated to the cultivation of information relevant to the team, and elimination of irrelevant data. Furthermore, this analyst can train team members to self-serve.
- Self service dashboards: Data analysts can train team members to operate a company’s metadata search engine, using metadata management products such as Google’s Data Catalog or the more popular self-service analytics tool, Power BI from Microsoft. Many platforms are designed to aggregate data into intuitive reports on the fly, making information even more accessible to end users.
- Clearly defined metadata: Healthy, well-structured metadata makes for better search results.
Culture is Key
You can pick all the right tools, implement the perfect system, and train employees to use those tools, but data democratization best shines when employees are primed for independence, self-motivated, and understand the value of the tools at their disposal.
Read next: Top 9 Data Modeling Tools & Software 2021