Data Transformation and Governance at Bouygues Telecom

Read more below on how I set up this data transformation and governance exercise with Tableau:

The starting point: The challenges of being a newcomer

When I joined my business unit at Bouygues Telecom in 2020, I encountered difficulties in understanding the existing dashboards. I faced four main challenges:

• As the telecoms sector was a new industry to me, I did not understand all the technical vocabulary and acronyms that were used in the dashboard
• My business unit had hundreds of dashboards and some of them were duplicates or overlapping
• Documentation for self-service was limited. Thus, it was hard to figure out by myself what the most important KPIs were, how they were computed, and for which operational teams they were useful
• And specifically, as a data analyst, it was difficult to onboard newcomers since access rights were maintained manually.

I shared this feedback with my lead and several internal clients with different profiles (managers, operational teams, newcomers vs. historical clients), and it appeared that most of them shared my vision and difficulties. Thus, I was assigned to facilitate the understanding of, and access to, dashboards in my business unit a few weeks after starting my job at Bouygues Telecom.

The planning process: Start with the end in mind

To make this project successful, I had 3 main goals to achieve:

• Accelerate business decisions by reducing the number of dashboards
• Up-level stakeholder understanding by providing better documentation
• Automate access rights to save time

I included several groups of stakeholders in this project including fellow data analysts, my internal clients, and Bouygues Telecom’s Tableau centralized team so everyone was aligned and could share any feedback on my methodology.

Methodology: Begin and end with the direct users

My methodology was made up of four main steps:

• Confirming the main pain points with the direct users. To do so, I did a dozen interviews with my internal clients to identify their main points and the ideal solutions they would want to have implemented in Tableau
• Cleaning my business unit’s Tableau Server. This involved 3 main actions: Archiving all the obsolete dashboards that had not been viewed for the past 6 months, merging dashboards that had redundant information, and finally adding descriptions and keywords to the remaining dashboards
• Building a new organizational system. Using the descriptions, I was able to design a tree structure on Tableau Server that had only 2 levels of hierarchy and was organized by main topics and subtopics. For this new tree structure, I created a new access rights map as well as several Tableau user groups to define who can view what based on the identity of the user
• Preparing the transition. I carried out user tests with clients from different departments and equipped with different Tableau licenses, to ensure the transition ran very smoothly. Finally, I shared extensive documentation to prepare my clients for the change

Business impact: Very positive results were obtained after the completion of the project

From a quantitative perspective, we achieved the following results:

• The number of dashboards in my business unit was reduced by 25%
• Any Tableau dashboard can be accessed in less than two clicks with the new tree structure
• Every dashboard has been documented with useful descriptions
• The time to add new viewers was sped up by 70x

From a more qualitative approach, other great benefits were obtained:

• Internal clients are much more independent in finding information
• Security was reinforced with the new access rights and Tableau user filters