3 data governance approaches: which one is best for your business?

Data governance is an ongoing practice that ensures your data is accurate, consistent, and secure. It instills a sense of trust in the data you’re using.

This post will cover the different data governance approaches you can apply in your organization: conservative, liberal, or blended. You’ll understand the different levels of access with each approach, and how to manage varying roles and permissions within your team. By the end, you’ll know which approach to data governance is best for your business and how to implement it effectively. 

Let’s dive in.

1. The conservative approach 

The conservative approach is the strictest approach you can take to your data governance strategy. It restricts the majority of users to view or read-only permissions, with higher-level access reserved for more senior roles: team leads, dataset owners, or lead analysts who have demonstrated sufficient analytics knowledge.

Larger organizations and enterprises favor this approach because of its minimum risk. 

2. The liberal approach 

The liberal approach gives the majority of users permission to edit and access your dataset, whether that’s updating information or defining events, queries, and dashboards. The liberal approach is the most trusting approach, with organizations allowing members to largely self-govern and manage the shared dataset or spaces on their own.

This approach works best for smaller, leaner teams that err on the more technical side. Because all team members will be given analyst-level access, they must have previous experience with analytics tools and high data maturity.

3. The blended approach

The blended approach is a mix of conservative and liberal. It gives team leads access to shared data and definitions, which they can provide to their team on an as-needed basis. New team members will default to viewer access. 

The blended approach is the best option for growing teams and organizations who want to mitigate risk when it comes to their data governance practices, but also want to enable ownership and innovation across their team. 

Different roles and permissions for your data team

After reviewing your current governance practices, it’s time to reevaluate user permissions and roles. Before you do this, it’s important to make sure you understand what each role is and what permission level they should have within your account. Here are some key roles and descriptions: 

• Analysis admin: the most senior role. Has the largest range of access and can manage projects, billing info, single sign-on (SSO) settings, the home dashboard, and delete teammates or change their roles.

• Analytics admin: second to the analysis admin, this role has the permissions of an analyst and can manage verified events and snapshots, connect new warehouses and sources, and edit change history

• Frequent analyst: this role has the permissions of a consumer and can create events, categories, and segments, and update property notes

• Ad-hoc analyst: this role has limited analysis and dataset access, but more than read-only. They can manage reports and dashboards, as well as run reports and invite other teammates.

Once you understand the different roles, you can then reassign permissions and restrict event creation and modification permissions in your shared space. Doing so helps you maintain data trust across different teams and roles.

Here’s an example of what that could look like across your team:

How to divide permission levels across teams

Here’s a breakdown of how each data governance approach can be divided across your teams and different roles: 

Option 1: blended

• Analysis admin: 5%

• Analytics admin: 15%

• Frequent analyst: 30%

• Ad-hoc analyst: 40%

• Read-only: 10%

Option 2: conservative

• Analysis admin: 5%

• Analytics admin: 5%

• Frequent analyst: 10%

• Ad-hoc analyst: 40%

• Read-only: 40%

Option 3: liberal

• Analysis admin: 5%

• Analytics admin: 15%

• Frequent analyst: 65%

• Ad-hoc analyst: 10%

• Read-only: 5%

Note: in general, there should be fewer analysis admins than analytics admins, and fewer analytics admins than analysts. Unless your organization is very small, the majority of users should have analyst or read-only permissions.

What criteria will you use to decide what permission level users should have? Check out this breakdown to guide you:

Option 1: blended

• All team leads will be set up with analytics admin access by default

• All new users will be set up with ad-hoc analyst access by default, and may request an upgrade in permissions from the analysis admin, with manager approval

Option 2: conservative

• All team leads will be set up with analytics admin access by default

• Users who have completed your analytics platform and practices certification will be set up with frequent analyst access by default

• All new users will be set up with ad-hoc analyst access by default, and may request an upgrade in permissions from the analysis admin with manager approval (but will need to be trained before upgrading permission level)

Option 3: liberal

• All team leads will be set up with analytics admin access by default

• All new users will be set up with frequent analyst access by default, and may request an upgrade in permissions from the analysis admin with manager approval

Empower your teams with valuable insights

Choosing the right data governance approach for your organization is crucial to maximizing the value of your dataset, ensuring you can run analysis quickly and accurately across your entire organization. By providing the right people with the right level of access, you empower them to make data-driven decisions without compromising on security and compliance.