As the top online retailer in the Netherlands and Belgium, bol.com has grown massively in a short amount of time. As of 2020, the company serves 12 million active customers and offers over 30 million items. The bol.com team’s overloaded Hadoop-based analytics infrastructure was not delivering the performance needed to support their business intelligence program on Tableau. As they shifted analytics infrastructure to Google BigQuery, they chose AtScale as a semantic layer to ensure a live, fast connection to historical data without requiring complex data engineering. This helped the company accelerate its cloud adoption while saving money on Google BigQuery compute costs.
of Tableau workbooks loaded in less than 10 seconds
Provided a live connection to fresh and historical data for 1600 users
Reduced cost by 91% for potential savings of nearly $10 million
Breaking a data bottleneck with a new BI solution
As the top online retailer in the Netherlands and Belgium, bol.com has grown massively in a short amount of time with an innovative data-driven team. As the company scaled, the data began evaluating alternatives to their overloaded Hadoop cluster that was taking too long to run some jobs.
At the time, the company’s analysts were using Platfora for data preparation and visualization. Shortly after the go-live, Platfora announced its acquisition by Workday and with that the discontinuation of the product. With this as a catalyst, bol.com began looking for a new solution to support their BI and analytics program.
Self-sevice was a top priority for the bol.com team. As they looked for new technology partners, they wanted to integrate a semantic layer solution that could cover all data assets, now and in the future. Further, they wanted to ensure compatibility with whatever BI and analysis tools they may use in the future.
Applying a semantic layer to bring live updated view of key enterprise metrics
The team at bol.com captured the lessons from their Platfora implementation and completely renewed their BI technology stack. They moved all their data from the on-premises Hadoop cluster to Google BigQuery on the Google Cloud Platform. The company also moved its team to Tableau for visualizing data and off of Platfora.
By leveraging AtScale’s semantic layer solution between Google BigQuery and Tableau, the company provided its 1600 BI users with a live, fast connection to fresh and historical data without the need for complex data engineering.
As a result, bol.com’s BI users now have a live connection to data to perform more timely analysis. The bol.com semantic layer ensures accuracy and consistency for data analysis and reporting across the organization. This single version of the truth is fundamental for achieving a true self-service analytics culture.
$10 million dollars in savings and better, faster data
The move to Google Cloud provided a lot of computational power during the peak months, when bol.com is busiest with holiday shopping. By leveraging AtScale’s semantic layer, bol.com’s BI users have a live connection to data to perform more timely analysis to better work during this crucial season.
AtScale’s unique query acceleration capability leveraging intelligent aggregation are utilized in more than half of bol.com queries. In an internal analysis of performance, bol.com estimates that over 70% of their Tableau workbooks load in less than 10 seconds. These same acceleration structures drastically reduce the amount of compute required to run each analyst’s query, providing bol.com with additional ROI with Google BigQuery.
AtScale’s ability to intelligently manage performance and optimize cloud resource consumption reduced cost by 91% for a potential savings of nearly $10 million annually.
Semantic Layer, Analytics Modernization, Query Acceleration
Faster time to insights, integrated toolset, lower cost, faster and more consistent analytics
Key product components
Google BigQuery, Excel, Tableau