In 2021, PetCenter was going through a challenging period. The company, which operates an e-shop and more than 150 pet supply stores in the Czech Republic and Slovakia, was missing profits and lacked transparency in its inventory and sales data. Newly appointed crisis manager Robert Vojáček was tasked with reviving the company, so he turned to us for help. 

„PetCenter was running an e-commerce store and more than 100 stores, but it had no visibility into surpluses and shortages, what was selling and what was generating margin,” summarizes Robert Vojáček, PetCenter’s interim manager.

PetCenter was also struggling with outdated systems (most of the reporting was done in Excel), inefficient data integration and a lack of sales history. This hindered informed decision-making and company growth.

And why did Robert Vojáček choose Revolt BI?

„I had worked with Revolt BI before. I knew them from Zoot and Pietro Filipi. I knew they understood the situation and the needs of a crisis manager.”

What the client says

Our solution

Our response was fast and targeted. As Revolt BI emphasizes results, collaboration and transparency, we quickly mobilized resources and began the transformation of the PetCenter data ecosystem. We deployed a team of three experts who worked closely with the company and kicked off our collaboration with an intensive workshop. Our goals were clear: build a robust data pipeline from multiple sources, implement Keboola and Tableau, and upskill their internal team on Tableau. Our efforts culminated in the creation of six comprehensive reports that provided the clarity and insight needed to address financial challenges.

PetCenter and Revolt BI worked together for two years, during which time there were significant improvements in PetCenter’s data ecosystem and approach.

How we built a data culture at PetCenter

Phase 1 – Getting to know the company better

The first step was understanding PetCenter’s unique needs through a workshop. As specialists in retail reporting solutions, we understand that each client brings a different set of challenges and requirements. The initial workshop, held at the client’s site, was used to develop a business data model (BDM) and logical data model (LDM). These formed the basis for the next phases of the project. Based on this analysis, Keboola and Tableau tools were implemented and PetCenter was migrated to the new ERP system – ESO9.

Workshop overview:

Our senior consultants communicated directly with PetCenter management to identify their reporting needs and goals. As a result, we were all on the same page, which set the foundation for a successful collaboration.

LDM model after 3-4 months:

PetCenter’s BI capabilities were primarily Excel-based, with historical data trapped in the outdated BarIS ERP system. The implementation of Keboola and Tableau allowed for the transition to the more robust ESO9 ERP system, which meant a complete system redesign.

Phase 2 – After the workshop

After the workshop, our team compiled the information gathered into a comprehensive plan, set clear priorities and assembled a bespoke team for PetCenter. This team included two data engineers, a project manager and a business consultant, while on the client side four users were trained to work directly with the data and create reports for a wider audience of 30-40 BI users.

Priority use case:

Our immediate goal was to deliver financial and performance reports to PetCenter’s senior management and investors. This key task was accomplished within just two months of our first engagement.

Phase 3 – Project Launch

We started the project as a Proof of Concept (PoC), which provided tangible results and laid the foundation for an agile approach based on sprints. A dedicated team of three Revolt BI experts worked closely with the customer, connecting data sources and creating six key reports. In addition, we tackled the challenges of categorizing data using data science, which enabled effective product categorization.

Phase 4 – Data discovery

During this phase, we linked the data into Tableau, enabling data analysis for 4 users and informed decision making for approximately 30-40 users who used the results of the reports.