“The impact [of Coalesce Catalog] is clear. Our weekly KPI review is now twice as efficient — down from two hours to one. We now spend less time challenging numbers and more time focusing on strategy and action.”
A data closet in disarray
Challenges
Rapid growth left 8,000 Snowflake tables with little governance
Over 600 dashboards caused redundancy and confusion
Team wasted time answering questions and reconciling conflicting numbers
Vestiaire Collective pulls in data from a variety of sources: operational data from backend systems, customer data from platforms such as Snowplow, Salesforce, and Zendesk, logistics data, payments, and more. A lean central data team led by Vira Douangphouxay, Director of Data Analytics, is responsible for ingesting and preparing the data for the analysts embedded across the various business units.
This decentralized model empowered the organization’s skilled analysts with easily accessible data, but also created sprawling assets with little to no governance in place. Within two years, the company had more than 8,000 tables in Snowflake and 600+ Tableau dashboards. The result was duplication, inconsistent answers, and mounting data chaos.
“Every day we found ourselves spending 3 to 4 hours just answering questions, trying to help people find the right table, the right data,” says Vira. “And even within our own team, we sometimes ended up having different answers because there were so many tables — it really affected our productivity.”
Vira knew that they needed a better solution to “clean out the closet” and get a handle on all their data.
Ready for the runway
Solution
Standardized on Snowflake early on as the unified data platform
Evaluated a number of different data catalog solutions
After successful POC, adopted Coalesce Catalog as discovery and governance platform
As soon as the company made the decision to move to Snowflake, it began building an agile, scalable data stack around it, including Airflow for orchestration and Tableau for data visualization. First and foremost, however, the team needed to select a data catalog. “It was a prerequisite for us, something we knew we needed to have based on best practices many of us had learned from previous companies,” Vira says.
At first they evaluated another popular data catalog that offered a large amount of features, and while it looked fancy, they worried it was too complex. “We were only just starting to implement a catalog, and our team was somewhat junior at the time. This other solution was too noisy and distracting for our needs,” he says.
That’s when Vestiaire Collective started to look at Coalesce Catalog as a possible fit. “The idea was to find a solution that would integrate with our data stack with minimal effort on our part,” Vira explains.
A single source of truth is always in fashion
Results
Data Analytics team slashed daily time spent answering questions from 3–4 hours down to just 1 hour
Recurring business review meeting shortened from 2 hours to 1, and analyst prep time cut by 50%
User engagement of Catalog has grown to 60 people across the organization, all of whom now contribute to documentation directly
Single source of truth has fostered newfound companywide confidence in data
With a data catalog now in place, the chaos the Data Analytics team was grappling with is a thing of the past, and they have come to rely on Catalog as the central place to quickly find information. “Instead of hunting through tables, I can just type in a keyword or metric and see a list of results with star ratings,” says Vira. “If a table has five stars, I know that’s the one to use.” He adds that being able to view usage and sample queries saves them a lot of time: “That’s the real joy of using a data catalog.”
Most importantly, Catalog is now central to Vestiaire’s strategy of developing a single source of truth for the entire organization, which Vira explains is built upon five core pillars: certified KPIs, high technical standards, documentation, certified data visualizations, and training.
Vira says that his team’s work is changing the culture of the company, and there’s a newfound confidence in the data people are seeing. “During a CEO business review to determine why GMV had dropped, our team was faced with conflicting data from various stakeholders,” he explains. “We presented our single-source-of-truth dashboard. Using Coalesce Catalog, we were able to show everyone the agreed-upon definition of GMV and its complete data lineage in real time. This created an ‘aha’ moment that aligned the entire room and made the value of a single source of truth undeniable.”
As Vira notes, “The impact is clear. Our weekly KPI review is now twice as efficient — down from two hours to one. We now spend less time challenging numbers and more time focusing on strategy and action.”
Catalog helps provide the data granularity needed to investigate problems accurately, without wasting time on ad hoc analyses that risk leading to the wrong conclusions. “We are trying to enable and empower as many people as possible in order to scale the company. We’re building a platform for people to do their jobs more easily, whether that’s running analyses or building reports,” he says.
While initially it was only Vira’s team of analytics engineers who were using Catalog, he says that number has grown thanks to Catalog’s ease of use. “Today, there are as many as 60 users, including people on the product team and other business teams,” he says. “And they’re not just using it — they’re contributing to the actual development of our data catalog.” Vira explains that this is part of the reason Vestiaire decentralized its analytics — because business analysts are closer to the business needs, they often better understand the context of the data they are looking at: “That’s why we’ve given them the permissions to edit and create documentation themselves.”
Thanks to Catalog, the three to four hours his team spent every day answering questions has dropped to just about one hour per day. “And now these questions are no longer basic ones such as ‘Where can I find my data?’, but rather more advanced questions that do require our expertise to answer. So Catalog has been a huge boost to our team’s productivity.”
The team has started using the platform’s AI features, and is strategizing how to leverage them in exciting new initiatives going forward. Even though the amount of questions the team is regularly asked has been greatly reduced thanks to Catalog, Vira would like the platform’s AI assistant to automatically answer the remainder that still come in via Slack: “Ideally, Catalog itself will handle those questions for us.”
Vira says his team is also investigating how to develop analyst copilots, which could eventually take over some tasks analysts handle today, such as crunching data or writing queries. “All of the tools we will use to build this require a solid data foundation, which is why the work we’re doing with Catalog today is so important,” he says. “This foundation is what’s called a semantic layer, where all the documentation and definitions live. For us, this layer is critical — it’s the base we’ll use for every AI copilot we build moving forward.”