With the 4th of July celebration quickly approaching, I thought it would be a fun time to talk about democracy in the context of data, specifically data management and the newish cry for the “democratization” of data.
As I reflected on what it means to be a true democracy with its benefits and challenges, and the nature of our country as more of a federal republic versus a pure democracy, I wondered, “Are there some lessons learned from the political philosophy realm that we can apply to the concept of democratization of data?”
The answer is yes. In this article, you’ll learn what democratizing your data means, the value it can provide to your organization, and the benefits and pitfalls.
We’ll take some time to think about concepts like voting, egalitarianism, pluralism, trust, education, literacy, policies, freedom, and a central governing body that has a sort of “final say” when it comes to data decisions that affect the larger enterprise.
If nothing else, in the spirit of Independence Day, it’s probably your patriotic duty to read on.
Definitions, Definitions, Definitions
Let’s start with some fundamental definitions for the layperson. Instead of using PolySci 400-level course definitions, I’ve chosen some common-sense explanations that are high-level enough that we can also apply them to data management.
Democracy: Giving the people direct access to have a voice in the creation, application, and overall direction of the larger entity to drive towards a shared (or even at times, disjointed) vision. In our case, let’s just say data (and thus power) to the people!
I’ve put all the other terms below, so please review those definitions before you read on. Don’t skip this part because it’s important.
Ok, having read the definitions, you’ve seen that I resisted the urge to paint each one in the context of data management, but let’s do some of that now.
Democratization and Data Management
Looking back, in the sixth century B.C., for Athenian democracy to work appropriately, it helped that there was a small number of citizens, and they were all educated.
In our modern American democracy, we operate in more of a federated republic with a much larger population with varying interests, education levels, and backgrounds such that pluralism (a political system with more than one center of power) is going to exist.
The same can be said of larger institutions when it comes to data management, particularly at the enterprise level. There will be various constituents based on department, function, level, interests, education, and analytics maturity.
If we are to think about “data to the people” in this context, we would definitely want to have some discussions around organizational change management to help ensure an educated population, a set of consistent policies to help guide the distributed creation and curation of data assets, and some sort of governance to help prevent complete chaos. I like to call this responsible data asset development.
Data Democratization: What and Why
So “data to the people,” or the “democratization of data” as it relates to data management, refers to the process of supplying all the right data to the “citizens” in your organization in the right form when they need it.
That’s easier said than done. How do you prevent chaos when data is pouring in at an exponential rate, the business is asking for new insights, IT can’t keep up, and unsanctioned assets are being developed at warp speed by people using next-generation business intelligence tools designed for the non-technical user?
The solution is to create a balance between governance and democratization of data.
Advantages to Democratizing Your Data
There are some clear benefits to the democratization of data. In theory:
- The data is being made accessible to those who know it best.
- Democratizing data asset development and creation is a force multiplier in that capacity is no longer throttled by the number of hours IT has available; this should produce a quicker time to value by minimizing this reliance.
- New levels of data sharing and awareness can occur as communities form and the social fabric associated with the enablement of democratization of data develops. This can lead to a better shared understanding and use of data, while also improving the dialogue around how best to deal with challenges inherent in the data.
So faster time to value, better awareness, and the establishment of a civil society to help with a shared dialogue can all improve an organization and their management of data. When these a occur, a true data marketplace can be developed to allow “citizens” to discover data assets, make social comments regarding the data assets quality, usefulness, meaningfulness (think Yelp for data), and even crowd source business data in a shared business glossary.
Issues with Democratizing Your Data
Democratization sounds like a great idea, but when is it not? Here are some pitfalls to consider:
- The uncontrolled proliferation of data assets.
- Duplication and near-duplication of data assets.
- Confusion over terms – fields called the same thing in one data asset do not have the same meaning in another.
- Creation of dark assets unknown to all but one person or a few.
- Reporting based on non-sanctioned assets – ungoverned assets used for reporting can put an organization at risk, particularly if those numbers are reported to the street.
- Time wasted reconciling numbers from different assets that should tie out.
- Not having a strong enough governing body or set of policies won’t provide the set of checks and balances that democracies need.
- Poor overall management or tooling can make it difficult to know and understand asset lineage.
- Depending on technical platform, uncontrolled asset creation and exploration can have real cost implications for data work in the cloud as compute costs can get out of control without proper monitoring and guidance.
Is it worth it?
There is clearly both an upside and downside to data democratization. The answer of whether it’s worth it to your organization should center around this idea that too much of anything is rarely good.
Striking the balance between full-out democratization and providing the right level of governance can hopefully put your organization in a position of moderation where many of the benefits of data democratization can be realized while avoiding or mitigating many of the risks.
You want your citizens to have freedom, but you also must ensure they are properly educated, have the access they need to that education so that policies can be appreciated and understood, and a community is fostered to build out the social fabric to continue to benefit the organization as a whole.
Moving beyond this July 4th weekend, reflect on the benefits and pitfalls of democratized data management, and start looking for opportunities to move your organization in the right direction. As you make the shift, Onebridge can help with the organizational change management, technology enablement, and process considerations to help you on your data democratization journey.
Have more questions about data democratization? Contact us