Democratized or Centralized - choosing your workflow

Last modified: November 04, 2019 • Reading Time: 5 minutes

Contribute to Data School

We are actively working on this chapter. We are looking for two types of contributions:

  1. Help writing this chapter
  2. Share your story about working through this problem at work

Please reach out to @Matt David on our slack to discuss what you want to contribute.

At this stage you need to make a big choice in your organization - who do you want to be making the charts and dashboards? Do you want a centralized workflow where all of the charts and dashboards are created by a core team, or do you want to be largely democratized where business users are enabled to self-serve and create their own charts?

The choice here has dramatic impacts on the business, and while we lean strongly toward the benefits of a democratized workflow we realize it’s not always right for every business or every stage.

Here we attempt to clearly articulate the pros and cons of both workflows. The choice here is painted fairly black and white because that’s just the reality of the implementation. If you choose to stay centralized and your central data team is less than around 50 people, there’s little need to do this final Mart stage of sophistication. It is more work to build and maintain, and if the only people using the data are full-time professionals then their need for this slicing is minimal.

The choice can impact what BI products you use, and completely changes how you organize your teams, trainings, and even who you hire.

Centralized data

In Progress

This has been the traditional approach, more due to database and product limitations than because it’s the ideal way to do things. Database storage and compute resources used to be very constrained, but now these resources are cheap. Products used to require advanced skills to analyze data, but now there are drag and drop interfaces to find insights.

With these changes in mind, let’s review the current set of pros and cons for centralized data access.

Pros

  • Less errors
  • Less chance of PII being handled incorrectly

Cons

  • Central team bombarded with requests
  • Curiosity has a gatekeeper
  • Hard to scale

Democratized data

In Progress

This is a new approach to working with data in an organization. It comes from a belief in employees’ abilities to answer their own questions if given the ability. While this method takes some up front work moving from centralized to democratized to get data in a good shape, this approach can scale as your company grows.

Let’s explore why democratized data is so powerful and acknowledge some of the drawbacks that keep other teams centralized.

Pros

  • More analysis
  • Empowered employees

Cons

  • More chance for errors
  • If data is not properly sanitized, PII could be handled incorrectly

Written by: Dave Fowler
Reviewed by: Matt David

Next – Why Build Data Marts

Get new data chapters sent right to your Inbox