This is the first episode of the second season of the Zeenea Effective Data Governance Framework series.
Divided into three parts, this second part will focus on Adaptation. This consists of :
- Organizing your Data Office
- Building a data community
- Creating Data Awareness
For this first episode, we will give you the keys to build your data personas in order to set up a clear and well defined Data Office.
Season 1: Alignment
- RUnderstand the context
- RGet the right people
- RPrepare for action
Evaluate your Data maturity
Specify your Data strategy
Build a SWOT analysis
Season 2: Adapting
- RCreate your personas
- RIdentify key roles
- RSet your objectives
Organize your Data Office
Organize your Data Community
Creating Data Awareness
Season 3: Implementing Metadata Management with a Data Catalog
- RGet to know your data
- RIterate your data catalog
The importance of metadata
6 weeks to start your data governance journey
In the first season, we shared our best practices to help you align your data strategy with your company. For us, it is essential to:
- Assess the maturity of your data
- Specify your Data Strategy by building OKRs,
- Get sponsorship,
- Build an effective SWOT analysis.
In this first episode, we will teach you how to build your Data Office.
The evolution of Data Offices in companies
At Zeenea, we believe in Agile Data Governance.
Previous implementations of data governance within organizations have rarely been successful. The Data Office often focuses too much on technical management or a strict control of data.
For data users who strive to experiment and innovate around data, Data Office behavior is often synonymous with restrictions, limitations, and cumbersome bureaucracy.
Some will have gloomy visions of data locked up in dark catacombs, only accessible after months of administrative hassle. Others will recall the wasted energy at meetings, updating spreadsheets and maintaining wikis, only to find that no one was ever benefiting from the fruits of their labor.
Companies today are conditioned by regulatory compliance to guarantee data privacy, data security, and to ensure risk management.
That said, taking a more offensive approach towards improving the use of data in an organization by making sure the data is useful, usable and exploited is a crucial undertaking.
Using modern organizational paradigms with new ways of interacting is a good way to set up an efficient Data Office flat organisation.
Below are the typical roles of a Data Office, although very often, some roles are carried out by the same person:
- Chief data officer
- Data related Portfolio/Program/Project managers
- Data Engineers / Architects
- Data scientists
- Data analysts
- Data Stewards
Creating data personas
An efficient way of specifying the roles of Data Office stakeholders is to work on their personas.
By conducting one on one interviews, you will learn a lot about them: context, goals and expectations. The OKRs map is a good guide for building those by asking accurate questions.
Here is an example of a persona template:
Some useful tips:
- Personas should be displayed in the office of all Data Office team members.
- Make it fun, choose an avatar or a photo for each team member, write a small personal and professional bio, list their intrinsic values and work on the look and feel.
- Build one persona for each person, don’t build personas for teams
- Be very precise in the personas definition interviews, rephrase if necessary.
- Treat people with respect and consider all ideas equally.
- Print them and put them on the office walls for all team members to see.
Building cross functional teams
In order to get rid of Data and organisational silos, we recommend you organise your Data Office in Feature Teams (see literature on the Spotify feature teams framework on the internet).
The idea is to build cross functional teams to address a specific feature expected by your company.
The Spotify model defines the following teams:
Squads are cross-functional, autonomous teams that focus on one feature area. Each Squad has a unique mission that guides the work they do.
In season 1, episode 2, in our OKRs example, the CEO has 3 OKRs and the first OKR (Increase online sales by 2%) has generated 2 OKRs:
- Get the Data Lake ready for growth, handled by the CIO
- Get the data governed for growth, handled by the CDO.
There would then be 2 squads:
- Feature 1: get the Data Lake ready for growth
- Feature 2: get data governed for growth.
At the level below, multiple Squads coordinate within each other on the same feature area. They form a Tribe. Tribes help build alignment across Squads. Each Tribe has a Tribe Leader who is responsible for helping coordinate across Squads and encouraging collaboration.
In our example, for the Squad in charge of the feature “Get Data Governed for growth”, our OKRs map tells us that there is a Tribe in charge of “Get the Data Catalog ready”.
Even though Squads are autonomous, it’s important that specialists (Data Stewards, Analysts) align on best practices. Chapters are the family that each specialist has, helping to keep standards in place across a discipline.
Team members who are passionate about a topic can form a Guild, which essentially is a community of interest (for example: data quality). Anyone can join a Guild and they are completely voluntary. Whereas Chapters belong to a Tribe, Guilds can span different Tribes. There is no formal leader of a Guild. Rather, someone raises their hand to be the Guild Coordinator and help bring people together.
Here is an example of a Feature Team organization:
Don’t miss next week’s SE02 E01:
Building your Data Community, where we will help you adapt your organisation in order to become more data-driven.
Copyright Zeenea 2021, all rights reserved.