30% Cost Reduction for Loop Returns on Google Cloud

Loop Returns


Loop Returns

Tech stack

Google Cloud + dbt


Enterprise Analytics



Loop Returns helps the fastest-scaling e-commerce brands on Shopify to automate their return process and guide customers towards exchanges. Loop Returns uses Looker to monitor several key metrics, including e-commerce return volume, with over 100 dashboards. Datatonic worked with Loop Returns to optimise its Looker instance and rebuild dashboards to reduce unnecessary cloud spend and enable Loop Returns to use Looker efficiently and at scale as the business grows.

Our impact


  • Reduced total Google Cloud costs by 30%, enabling increased spending on innovation and research + development
  • Optimised data integration sync frequency to reduce the amount of manual overhead required 
  • Reduced Looker load times and number of errors by rebuilding Loop Returns’ five most used dashboards with Looker best practices


The challenge

Loop Returns needed help rebuilding its Looker instance with a focus on best practices. It wanted to improve the levels of consistency and efficiency compared to its original Looker instance. 75% of its Looker dashboards were unused at the time, and Loop Returns wanted to make all of its dashboards accessible and useful to reduce this.

This included the need to implement best practices, such as centralising transformation logic and leveraging persistent derived tables. Furthermore, Loop Returns had realised that its Looker costs were increasing with little reward and needed to make its use of Looker across the business more scalable and efficient.


Our solution

Datatonic conducted a detailed audit of Looker and BigQuery usage for Loop Returns, found improvement and cost-saving areas, and rebuilt the business’ Looker dashboards and charts. This involved: 

  • Rebuilding Loop Returns’ five most commonly used dashboards in Looker with best practices in mind
  • Upskilling the Loop Returns team and creating best practices documentation for future reference
  • Integrating dbt to the data stack so that the business could build easier-to-maintain models outside of LookML
  • Optimised BigQuery spending using reservations and Flex Slots, reducing total spending by 30%
  • Migrating several models into dbt to rebuild critical dashboards
  • Implementing design best practices to minimise broken content downstream


With dbt + an optimised Looker instance, Loop Returns is now able to run better tests on its models and reduce the amount of broken content downstream. By depreciating unused Looker content and optimising its BigQuery pricing, Loop Returns drastically reduced its spending on data and BI.

Lastly, rebuilding its most-used dashboards and providing additional Looker training and documentation enabled Loop Returns to maximise reusability and scalability across its Looker instance.