Difficulties and Challenges of Data Democratization
There is perhaps nothing more vexing than a very powerful person with a strong need for their “heroic narrative” to come true. You’ve probably seen this in your organization. Data democratization becomes very difficult with this type of ethos embedded in the organization. Indeed, any sort of process that you might want to implement with data democratization can almost be rendered moot by a top executive with such a mindset. Sometimes rivalry is at the root of data democratization problems. Many organizations have rivalries of all sorts going on. Rivalries come about when:
Two or more units do the same or similar thing. Those units will start to compete with each other, and thus they start to hide things from each other in order to compete.
Two or more units (or leaders) might not be directly competing against each other, but those units (leaders) are still competing for resources or success.
To really get data democratized, you must look at these rivalries in your organization. Recognizing that your organization has a data democratization problem involves noticing symptoms like the following:
“The culture here makes it difficult.”
“We can’t let him/her/them see this or have access to that.”
“We have to tell the story softly to not offend.”
“We really can’t do that because we have so many ways to define X.”
“We need to change the data in the data warehouse.”
“Central IT just doesn’t understand the business.”
“Our data has quality problems.”
“We just need the data in our own systems. We can’t give it to another unit to manage.”
“Our analysts prefer their approach. They won’t want to participate with you.”
Some of these symptoms arise from basic human nature. So if we’re going to enter the business of data democratization, we have to understand that certain things are just ”the way they are.” In fact, information warfare is common across all animal species. Animals routinely manipulate information to gain a competitive advantage. Individuals in organizations do the same; we fashion information to meet our needs or purposes.
The human brain has evolved to process more information, particularly social information. When you combine information with the human capacity for very sophisticated social structure management, it leads to the natural instinct that “I need to use information to protect this part of my social structure.” We often refer to the phrase “information is power,” and that creates all sorts of predictable dysfunction (the symptoms shared above). It creates units inside the organization that want to hide information from perceived internal competitors, creating fragmented data systems. Sometimes deliberately not fixing data issues becomes an advantage for some units (fixing data might give other units an unwanted advantage). Along with dysfunction comes stress, and under stress, executive teams frustrated with the status quo of bad data quality, siloed systems, and an inability to make progress make imperfect decisions. When the going gets tough, the tough go shopping, and then they make bad decisions and buy the wrong thing (or do the wrong thing). This creates a cascading sequence of negative events.
Difficulties with IT Regarding Democratization
There is always a technical component to data democratization as well, and sometimes IT people can be part of the problem. There are some fascinating new technologies in data and analytics, and over the last five years or so, there has been a complete sea change in how we manage and analyze data. For instance, new systems can completely change the paradigm for how to perform data warehousing. As a result, everything that one knows about older ways of data warehousing can get in the way of adopting new approaches. Attachment to old systems — and the pride around building them — can hurt an organization.
So I tell all my IT people that I want “pirates.” Pirates are less loyal to the status quo. They may drop a tool (or technique) at the first sign of trouble in order to solve the next problem. I want extremely open-minded IT people who can be flexible and adjust their approach. Old-school data warehousing techniques and experts are at a big disadvantage. I admit, I put myself in this category. When I started with the new, next generation of analytics, everything I knew in the past was not helpful. I spent six months reeducating myself and my instincts to free myself from prior ways of doing things. This mental inertia about current technology and how new technology works plagues IT everywhere. As IT leaders, we must work on pushing through it.
[For more from the author on this topic, see “The Road to Data Democratization.”]
More Articles Like This
- Value-Driven Data Warehousing: Applying Agile Strategies to Increase Information Quality and Use
- Taming the Data Beast: Inheriting Key Insights from Data Warehousing Practice
- Challenges of the Current and Future State of Data Governance
- Well, Aren't You Smart: Culture and Skills for Information Superiority
- Semantic Ontologies: Be the Shepherd, Not the Sheep