Car aficionados are an enthusiastic bunch. Here’s how to tell whether you’re one of them:
- If there is a car show in town, you’re there.
- You can’t wait for your next issue of Car and Driver, Road & Track, or Autoweek to arrive.
- You spend countless hours restoring a beloved old car.
- When your car needs a new part, you make sure that it’s a genuine part branded by the manufacturer.
- You obsessively check car part inventories for the models you love.
It’s that fifth criterion that FCP Euro Auto Parts has focused on for 40 years. The Millford, Conn.-based company, which started as a brick-and-mortar seller of European auto parts, now boasts a vibrant web presence, including a web catalog with more than 150,000 parts.
Despite FCP Euro’s impressive growth and strong online presence, its founders saw data-related problems develop over time. The company had multiple source systems with varying levels of detail, as well as scattered data repositories, which made it difficult to locate data quickly and ensure its accuracy and reliability.
“Users could find answers, but either they were partial answers that didn’t give a complete picture of the business or they were competing answers,” explained Brandon Harris, who became the company’s first vice president of analytics last year. “In other words, there was a disconnect between how we painted the picture of the business and what the business was actually doing.”
Data Management Woes: Fragmentation, Integration, and Lack of Automation
Harris had known about the data management issues before joining FCP Euro. The company had at least eight data sources from different systems, each with a different level of detail. For example, some systems organized data by order number, while others by line item, customer ID, or email address. This resulted in fragmentation. Furthermore, FCP Euro’s ERP system, inventory management system, Zendesk service platform, and back-office system weren’t adequately integrated.
Before Harris arrived, FCP Euro attempted to solve its data management problem by building a business intelligence platform. However, that platform didn’t work out as expected. Harris surmised that while company leaders were excellent businesspeople and car enthusiasts, they lacked the technical expertise to maintain and update the platform.
Harris said that the main issue was that there was no automation to support data loading. Therefore, the only options were to use “a lot of duct-taped scripts to [ingest] data or people setting calendar reminders every Tuesday at 2 p.m. to export a CSV file from one location and upload to another.”
Data Management Overhaul in Four Steps
Before officially starting at FCP Euro in June 2022, Harris began suggesting data management improvements. He considered the company’s small IT staff and conservative IT budget and focused on making the data accessible and accurate for effective decision-making. To achieve this, he recommended automation of data ingestion, consolidation of all data, and new integrations.
FCP Euro’s first step was to find a better way to transfer data between different points. The existing data pipelines were prone to breaking and required the IT team to spend time fixing them. “I had seen from my prior roles that this could become a big time sink,” Harris noted.
Rather than investing in rebuilding the pipeline system or adopting an expensive platform, FCP Euro decided to implement Fivetran, an extract-and-load platform that pulls data from multiple sources. Fivetran has connectors that communicate with various APIs and endpoints, making it possible to extract information from almost any source.
After addressing the data transfer issue, Harris focused on securely consolidating and storing data. He chose Snowflake, a data cloud that allows for the efficient upload, organization, management, and storage of data. Harris believed that Snowflake would ensure that FCP Euro would not be hindered by growing pains as it expanded over time.
The third challenge in FCP Euro’s data management overhaul was to turn raw data into meaningful information that answers business questions. Harris recognized that this often required combining relevant pieces of data from different systems. He identified two potential solutions: developing an automated tool or finding an off-the-shelf platform that could handle this process.
The IT team was small, with only five members and limited coding experience. To avoid overburdening the team and to maximize their productivity, Harris chose to implement Coalesce. Coalesce is a drag-and-drop tool that allows users to define transformations and automatically generates Snowflake code to move and transform the data in the platform. The graphical nature of Coalesce enables users to visualize dependencies, such as the relationship between different tables.
With Coalesce, the team could now create graphs and documentation that provide a comprehensive view of a single topic, like current sales data, which it could add to the company’s internal Wiki. Harris explained that this approach enhanced understanding and trust within the team. When someone has a question, they can refer to the Wiki, where they can access relevant information.
The final step was to implement a tool that would present the results. Harris chose Sigma Computing’s tool, which allows for the rapid creation of interactive dashboards.
Benefits of FCP Euro’s New Data Management Tools
The capabilities provided by these new data management tools — corralling data in Fivetran, storing and managing data in Snowflake, combining data with Coalesce, and presenting the data with Sigma Computing – created a comprehensive reporting and analytics setup.
Gathering the right data sets for analysis became much easier. As an example, Harris compared how the old system and the new system worked in determining how a promotional code drove new customers and sales.
- The old system: To find out, the team would look at backend sales systems and order systems separately to locate all orders that used the promo code. They would also locate information about how much a customer purchased with the promo code and whether that customer was new. Certain data would remain inaccessible in the old system, such as how many customers put items in their cart but didn’t complete their purchase.
- The new system: The team can extract its Google Analytics data with Fivetran and put it into Snowflake along with sales data from its ERP system. The team can then use Coalesce to stitch the data together to identify common characteristics of customers who used the promo code but didn’t finish the transaction. “If we know the common characteristics of these users, we can go back to their customer profiles and get the complete picture,” Harris explained.
The suite of data management tools has proved useful in other business areas, as well. For instance, customer service agents can now access 360-degree customer profiles, as all customer touchpoints (e.g., customer interactions, return rate, historical spend, return rate) are linked together. This has allowed agents to obtain an accurate understanding of each customer by looking at the dashboard.
FCP Euro is also using the new technologies to improve inventory forecasting. By using the data from sales systems, ordering systems, and inventory systems, the company can identify historical purchasing patterns and determine the products to have in stock at the right time. “[Inventory forecasting] is much easier when you can get all the data and stitch it together,” Harris said.
The implementation of the new tools has resulted in benefits beyond data management. FCP Euro has saved 25% of the time needed for IT-related break/fix support, which has translated into savings. Furthermore, the cost of the new tools has been offset by decommissioning older, legacy platforms.
About the authorKaren D. Schwartz is a technology and business writer with more than 20 years of experience. She has written on a broad range of technology topics for publications including CIO, InformationWeek, GCN, FCW, FedTech, BizTech, eWeek and Government Executive.