I could not imagine that reducing the number of questions we got from stakeholders was possible before we had a data catalog
Unlocking Data Discovery
Started in 2009, Vestiaire Collective is the leading global online marketplace for desirable pre-loved fashion.
The company serves 23 million members across 78 countries and offers an unparalleled catalog of 5 million luxury and desirable fashion items with 6,000 brands available and 20,000 products added daily.
Jimmy Pang, Vestiaire Collective’s Business Intelligence Lead, came to Castor in 2021 to help him with his goals to manage data stakeholders and empower analysts to conduct their own analyses to deliver actionable insights.
Specifically, Jimmy’s mission was to “Empower stakeholders in business, product and tech to make their own decisions with data.”
When we first interacted with Jimmy and Vestiaire Collective one year ago, the company was looking to tackle this mission by solving three major challenges:
1. Documentation was manual, tedious, and dispersed across their stack including Snowflake for storage and compute, dbt for transformation, and Tableau for visualization.
2. Employees did not understand and trust the data. The BI team would spend a lot of time answering very basic data-related questions.
3. Onboardings were frequent and took a lot of time and effort.
Challenge 1: Tedious documentation
At Vestiaire Collective, table documentation is populated by the BI team. Consumers of this documentation are the business and product departments, plus some tech readership as well.
“Our biggest Castor users are business analysts,” Jimmy revealed. “When they are presented with new data, they are usually unsure of what it does and what every column means. For example, when there is a flag indicating that a column is valid, they use the catalog to understand exactly what is meant by ‘valid’ by checking the definition.”
Before Castor, the BI team was taking care of all the documentation in Confluence. The process was tedious and manual. Data people would create one page per schema in Confluence and then a list of documentation per table. This process was not only monotonous, but it was also more prone to human error than an automated way of documenting data.
Vestiaire Collective's documentation issue also stemmed from the difficulty to track ownership across the data ecosystem. Lack of ownership led to poor data assets maintenance and overall quality and trust decrease.
Solution: Automated documentation.
“Once we had plugged Castor in, organizing our knowledge became much easier. Both in Snowflake and Tableau,” Jimmy Pang, Business Intelligence Lead, Vestiaire Collective.
Castor automatically documents a large percentage of the company’s data warehouses, which reduces room for human errors. For the “last mile” documentation, the BI team documents tables manually. Castor then leverages its data lineage technology to propagate definitions on relevant columns.
Castor also allowed Vestiaire Collective to assign and track ownership of data assets. Assigning responsibility for data assets has proved efficient in incentivizing employees to document tables.
Challenge 2: Inefficiencies due to lack of context
Before Castor, stakeholders had no easy, straightforward place to look for context, such as column definitions other than just asking their colleagues.
As a result, the BI team would spend a lot of time answering fundamental questions linked to data assets. This time could have been used to focus on more business-critical topics.
“When you are a 100-person company, you can just tap on your colleagues’ shoulders to get information. After that, these kinds of practices are outdated, inefficient, and dangerous,” Jimmy said.
Solution: Enhancing data with rich context
“I could not imagine reducing the number of questions we got from stakeholders was possible before we had a data catalog” - Jimmy Pang, Business Intelligence Lead, Vestiaire Collective.
Castor drastically reduced the amount of time spent answering ad-hoc questions, by empowering stakeholders with a rich context and understanding of data assets.
Jimmy noted that Castor provided context around data assets that even data producers would never produce themselves.
Data producers could not always provide all the necessary context because they either don’t realize that specific context will be valuable, or they encounter technical barriers (regarding data lineage, for example). With Castor, the team could benefit from enriched data sets with data types and data lineage which turn out to be of utmost importance to data consumers who are trying to understand a data asset.
Jimmy explained that Castor allowed to free up a lot of time for the data analytics team to focus on more value-generating actions. For example, the team now has more time to focus on improving efficiency and the value of their data products.
Challenge 3: Regular onboardings
Vestiaire Collective is a fast-growing scale-up. The company is adding new employees every week. A key challenge was to enable these new employees with data.
Without Castor, new employees were struggling to become operational with the data. That is, they had problems when they needed to quickly understand data projects and their evolution, the data owned by the company, and where they could find it.
Without a catalog, new joiners were usually assigned a mentor to help get them running, take them through the current roadmap, the key people to contact, the tools used, metrics to track, and an overview of past achievements. This was a lengthy process and each onboarding ended up taking a few weeks.
Solution: A user-friendly data catalog.
“Castor has tremendously facilitated onboardings. I just onboarded a new employee last week and had nothing else to do than give her access to Castor.”- Jimmy Pang, Business Intelligence Lead, Vestiaire Collective.
With Castor’s data catalog, the on-boarded person now becomes fully independent and doesn't need a mentor to walk them through the company’s data assets. From column names definitions to metrics calculations, the new joiner can learn everything simply by scrolling through the catalog interface.
Giving people access to a rich catalog interface reduced Vestiaire Collective’s onboarding time from a few weeks to barely two days. It also suppressed the need to assign a mentor to each new employee.
One year after implementing Castor, Vestiaire Collective assessed that the tool has had a threefold impact.
- Castor made documentation a smooth and automated process when it used to be a tedious, manual, and error-prone one.
- Castor enabled employees to find, trust and understand the data swiftly. This tremendously reduced the number of ad-hoc requests flowing through the company.
- Castor empowered new joiners to become fully independent during the onboarding process.
Vestiaire Collective estimates that across these three elements, they have seen a 20% increase in team productivity since implementing Castor.
Subscribe to the Newsletter
We write about all the processes involved when leveraging data assets: from the modern data stack to data teams composition, to data governance. Our blog covers the technical and the less technical aspects of creating tangible value from data.
At Castor, we are building a data documentation tool for the Notion, Figma, Slack generation.
Or data-wise for the Fivetran, Looker, Snowflake, DBT aficionados. We designed our catalog software to be easy to use, delightful and friendly.
Want to check it out? Reach out to us and we will show you a demo.
Read More Success Stories
The story of how Payfit used CastorDoc to improve documentation, optimize storage, and empower decision-making across teams.
Bridging the Gap Between Data and Decision-Making: How JW Player Used CastorDoc to Democratize Data Access Across the Organization
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, Head of Data, Printify