Thibaut Gadiolet

Homeserve Builds a Self-service Data Platform with CastorDoc and Keyrus

Discover how Homeserve France boosts data team productivity and builds a self-serve data culture with CastorDoc and Keyrus.

Homeserve Builds a Self-service Data Platform with CastorDoc and Keyrus

This customer Story is contributed by Thibaut Gadiolet, Chief Data Officer at Homeserve France. HomeServe has been a specialist in repairs, construction, and maintenance of home equipment in France for over 20 years. With more than 1 million clients and 190,000 interventions per year, HomeServe boasts a network of 4,000 professionals.

Introduction

“Our end goal is to make sure everyone in our organization, regardless of their role, can work with data on their own to make better decisions.” Thibaut Gadiolet, Chief Data Officer, Homeserve

I started my tenure at HomeServe eight years ago, initially focusing on cybersecurity. After two years, my responsibilities transitioned to leading a small team dedicated to the development of a modern data platform.

As we crafted this data platform, I advocated for a new department to centralize data expertise, which I now lead. Our team, separate from traditional IT structures, is a specialized data department. Our data team is divided into four expertise: Data Engineers, Analyst Scientists, Analytics Engineer,s and Data Governance. Our end goal is to make sure everyone in our organization, regardless of their role, can work with data on their own to make better decisions.

Each department, from finance to partnerships, has its unique focus, yet they can all leverage data for improved decision-making. For instance, our partnership team needs to self-generate precise reports for monthly touchpoints with key partners. In marketing, data enables tailored campaign selections and performance analyses. We want these team to be able to self-serve data instead of sending requests to the data team each time they have a data need.

In 2022, we began a collaboration with Keyrus, a consultancy expert in crafting data and digital technology strategies for enhancing business performance. As part of this collaboration, Antoine Giraud from Keyrus became a part of the Homeserve data analytics team to assist in developing our self-service capabilities. Keyrus contributed by offering expertise in various areas and facilitating the adoption of CastorDoc.

In 2023, we selected CastorDoc as our platform for data governance and to promote self-service access to data. CastorDoc stood out as the ideal choice to realize our vision for self-service data access.

Challenge: Scaling the Data Team at Homeserve

“The absence of proper documentation and governance resulted in inconsistencies in company knowledge, with different departments calculating key metrics like conversion rates in varied ways.” Thibaut Gadiolet, Chief Data Officer, Homeserve

At HomeServe, we faced a significant challenge with scalability. Our data team, despite its size, was handling the entire data lifecycle, from gathering data to creating specific reports for various needs.

This approach worked initially but wasn't going to hold up as the company expanded. With an increasing appetite for data and a growing workforce, scaling the data team in line with business growth was impractical. We recognized the need to empower different departments to work with data independently to achieve the scalability we aimed for.

As our data team expanded, the onboarding process became increasingly drawn out, primarily because of insufficient documentation. Typically, it took a new team member about a week to fully acclimate to their specific area of expertise.

Maintaining documentation was another significant hurdle. The sheer volume of this task led to it being frequently overlooked. Documentation was manually handled, making it a particularly arduous process. Moreover, the absence of proper documentation and governance resulted in inconsistencies in company knowledge, with different departments calculating key metrics like conversion rates in varied ways.

Finally, our data team dedicated significant manual effort to data modeling, encompassing the writing, testing, and deployment of SQL scripts. This process was not only time-consuming but also challenging to scale, thereby highlighting efficiency issues within the team. The lack of mechanisms to effectively track changes, manage versions, or review code further compounded these challenges, delaying projects and complicating the discovery of data models and transformations.

The realization that our productivity was declining and instances of data misuse were leading to inaccurate reporting underscored the necessity for a tool. We understood the imperative need for a solution that could address these growing challenges and support our scaling efforts.

Solution: Empowering HomeServe Through the Citizen Data Analyst Model

“CastorDoc is central to our Data Academy, guiding business users through analytics insights and helping demystify data constructs, definitions, and KPIs. It underlies our mission to democratize data access, standing as the technological backbone of our vision.” - Thibaut Gadiolet, Chief Data Officer, Homeserve

To establish a self-service model, addressing internal challenges within the data team and broadening data platform access across the company was essential.

This led to the initiation of a project this year aimed at empowering every employee to become a data analyst, in what we term the 'Citizen Data Analyst”' model.

Three pillars support this model:

  • The self-service data platform, accessible to all employees.
  • Data preparedness, adapting data complexity to user expertise, from basic datasets to advanced analytics models.
  • The Data Academy, promoting data literacy with training and workshops.

This approach breaks down traditional departmental barriers, facilitating unrestricted data access to enhance data-driven decision-making across the organization.

Within our 'Citizen Data Analyst' initiative, we partnered with CastorDoc and Keyrus to enhance our data handling capabilities.

Antoine Giraud from Keyrus became part of our team, bringing in best practices for data modeling, including the implementation of Dataform as a key modeling tool. This initiative aimed to enhance the productivity of our analytics engineers. His role extended to training analytics engineers on effective modeling and sourcing techniques, improving our systems for better error detection and resolution, and working alongside the data governance team to unify our approach to metrics and KPIs.

In choosing a data catalog tool, we sought a solution that matched our scale, supported our self-service goals, and was accessible to non-technical users. CastorDoc emerged as the clear choice, aligning perfectly with our vision for self-service. This partnership has been instrumental, benefiting both our data and business teams and allowing us to influence the product's development. CastorDoc's integration with our Google infrastructure and its user-friendly features, like the Chrome extension, have significantly driven its adoption within our organization.

The implementation of CastorDoc quickly became integral to our data teams' daily operations. Meanwhile, our governance team is methodically building our data catalog, ensuring thorough documentation for business users. Beyond mere training on tool usage, we're establishing a comprehensive academy to deeply embed the self-service model into our culture. This ensures our practices remain agile and adaptable to evolving tools.

CastorDoc is central to our Data Academy, guiding business users through analytics insights and helping demystify data constructs, definitions, and KPIs. It underlies our mission to democratize data access, standing as the technological backbone of our vision.

Impact: Streamlining Efficiency, Paving the Way for Self-Service

“With CastorDoc's lineage feature, the process of impact analysis has been transformed, allowing us to complete evaluations in just a few minutes—a task that previously demanded several hours of detailed investigation” Thibaut Gadiolet, Chief Data Officer, Homeserve

Our partnership with CastorDoc and Keyrus has been transformative for Homeserve. Measuring the exact scale of the benefits is complex as they are often tangled in productivity and data discoverability narratives, but the benefits are undeniable.

CastorDoc has enhanced HomeServe's operations across three primary areas: data onboarding, impact analysis, and documentation automation.

  • Data Onboarding: The introduction of CastorDoc dramatically decreased the time needed for new team members to familiarize themselves with their specific data domains, cutting down from a week to less than three hours.
  • Impact Analysis: With CastorDoc's lineage feature, the process of impact analysis has been transformed, allowing us to complete evaluations in just a few minutes—a task that previously demanded several hours of detailed investigation. This efficiency gain is due to CastorDoc's ability to trace data dependencies and relationships quickly, providing immediate insights into how changes in one area might affect others.
  • Documentation Automation: CastorDoc automated what was once a cumbersome documentation process. Documentation, a task often overlooked due to its scale, is now efficiently managed and prioritized, providing a visibility we were lacking in the past.

But the benefits of CastorDoc extend beyond these three elements. CastorDoc is envisaged as the central access point for employees to manage data utilization effectively. This involves understanding data, ensuring its quality, and guiding stakeholders through reports to find specific metrics. With its integration into our BI tools, CastorDoc is set to enhance efficiency and provide visibility for our data strategy, aiding all business users. We're particularly excited about features like Dashboard Q&A, enabling users to interact directly with the CastorDoc AI for instant dashboard insights.

Antoine from Keyrus has been crucial to this initiative's success. Keyrus has refined our data modeling processes and helped with documentation next practices. Antoine is working closely with our data governance team to help establish a definitive source for all data definitions, metrics, and KPIs. His expertise in data management and alignment has been invaluable, sharing best practices that have significantly improved our operations.

What’s ahead for Homeserve

As of now, Castor is primarily used by our data team, who are focusing on enriching the tool with information. This preparatory phase is crucial for its adoption by business teams. We're anticipating this transition to start happening as soon as next week when we launch with the partnerships team, bringing in a new set of business users to CastorDoc.

Our approach with CastorDoc is evolving. It's more than a governance tool; it's becoming an integrated part of our ecosystem. With new features like the Chrome extension and natural language querying, we're enhancing user experience, allowing stakeholders to leverage CastorDoc within their existing environment. Adoption is key, and we're fostering it by embedding CastorDoc into everyday workflows, making it an invisible yet indispensable part of data interaction. Although CastorDoc has already had a significant impact on the working of our data team; it is this strategy that will drive true value for our enterprise.

Moving forward, CastorDoc and Keyrus remain central to our mission—not only in streamlining data team operations and establishing robust processes but also in realizing a comprehensive, automated self-service model. This vision empowers every department to independently harness data, fostering an environment where data-driven decisions are the norm.

About

Get in Touch to Learn More

See Why Users Love CastorDoc
Fantastic tool for data discovery and documentation

“[I like] The easy to use interface and the speed of finding the relevant assets that you're looking for in your database. I also really enjoy the score given to each table, [which] lets you prioritize the results of your queries by how often certain data is used.” - Michal P., Head of Data