Skip to main content

Meeting the needs of online food consumers

Jeongyookgak relies on data insights to improve food supply chain operations on Databricks


Data volume processed with one-third of the engineering staff


Increase in the number of active customers with Databricks-powered CRM campaigns


More returns on push marketing investment

CLOUD: Azure

“With the Databricks Data Intelligence Platform, Jeongyookgak and Chorocmaeul have internalized their data engineering and analytics capabilities, moving from a traditional data mart with a fixed annual maintenance fee to a pay-as-you-go data intelligence platform that delivers data insights at a reasonable cost and beyond.”

— SooA Kwon, Data Leader, Jeongyookgak/Chorocmaeul

Jeongyookgak has achieved rapid annual growth while revolutionizing the distribution culture of the livestock industry by establishing a new food paradigm. In 2022, it acquired Chorocmaeul, the first eco-friendly direct to consumer (D2C) food franchise in Korea, which currently provides food service to 2.8 million customers through 360 stores nationwide. Using data insights, they seek to ensure product operations can support fast-changing food consumption trends and tap into online market opportunities. To resolve the data silos created by their legacy data infrastructure and democratize their data environment to all Chorocmaeul employees, Jeongyookgak switched to the Databricks Data Intelligence Platform. Now they can leverage real-time exploratory analytics to accelerate business growth through innovations in personalized customer experiences, targeted promotional sales and improved operational efficiency.

Low flexibility and scalability of the existing data system

The food and agricultural industry is undergoing a transformative shift, harnessing data to revolutionize operations, optimize distribution channels and enhance food quality across the entire supply chain. Despite Chorocmaeul's extensive experience in eco-friendly food distribution, their Oracle cloud-based data mart struggled to meet real-time demands. It only provided basic management indicators for the distribution business, such as trade volume, customer count and sales figures. It also failed to dynamically adjust performance indicators, such as tracking member counts for specific promotions and settling benefit costs. The fragmented nature of their data storage and analysis systems led to the inconvenience of using separate analysis environments after extracting source data. Sooa Kwon, Data Leader at Jeongyookgak/Chorocmaeul, noted, "The existing ETL environment lacked scalability and flexibility, making it difficult to perform exploratory analysis using source data."

Given the obstacles to real-time customer management and product status management due to data processing delays, Chorocmaeul also faced challenges in fresh ingredients distribution. It took approximately three days from data request to reception because data could only be transferred from the source by IT personnel. There was also an additional cost for applying data insight to service because analysis and model development had to be conducted manually in a separate analysis environment.

While planning and launching new services following their acquisition by Jeongyookgak, Chorocmaeul was in prime position to tap into a wider range of data resources to fuel innovation. To support emerging functions and services, data scattered across various cloud environments needed to be consolidated into a unified multicloud platform. In response, Chorocmaeul opted to establish the Databricks Data Intelligence Platform as the foundational backbone for their data analytics requirements.

Exploratory data analysis achieved with a centralized management system

With the Databricks Data Intelligence Platform, Chorocmaeul can now conduct real-time performance status analysis. For example, they can now monitor the performance of their new infant and toddler brand, Choroc Bebe, on a daily basis. This includes tracking metrics such as the number of members, sales and cost of benefits. They’re also able to promptly review customer feedback, observe trends in active user growth and analyze product sales quantity. These capabilities enable swift and accurate business decisions. Rapid analysis of sales volumes and supply trends allows them to extract insights to improve inventory management and predict demand more effectively. And they can conveniently secure additional products based on demand and address issues such as out-of-stock items.

Scalability is one of the key benefits delivered by the Data Intelligence Platform, as Chorocmaeul can now simultaneously analyze large-scale data on around 1,000 stores and approximately 30,000 product SKUs accumulated over 12 years since digitization. For example, through a customer churn prediction model that analyzes customer purchase cycles, they can build a personalized customer management strategy while also starting a CRM campaign to send automated messages during the period of customer churn risk. Additionally, data analysts can leverage Python and Apache Spark™ engines to conduct exploratory analyses across diverse domains such as supply chain and pricing policy.

Through the unified nature of the Databricks Data Intelligence Platform, various source data, such as Azure Cosmos DB, GCP Firebase and Oracle Cloud Infrastructure (OCI), have been centralized. This facilitates more efficient support for new app functions and services through integrated data management and sharing. Sooa Kwon elaborated, “Databricks delivers the benefits of efficient time, workforce and economical use while ensuring high flexibility and scalability for data management.” The adoption of Databricks by the Chorocmaeul team has been notable because the platform now enables direct data checking using SQL, data visualization, dashboard configuration with BI tools and data monitoring through Superset instead of solely relying solely on predesigned indicators. This shift has been highly advantageous in fostering a democratized and data-oriented corporate culture.

Upgrading customer management strategies with the Data Intelligence Platform

Chorocmaeul has upgraded its data engineering and analysis capabilities through their implementation of the Databricks Data Intelligence Platform. This has streamlined data access for employees, significantly reducing the need for time-consuming tasks like basic data extraction and search while also improving employee productivity. Since adopting the Databricks Platform, Chorocmaeul has scaled up its data pipeline by at least 100 times and its data volume by 1,000 times. Despite this considerable increase, the company has maintained its engineering personnel at around one-third of the previous size, with only a two- to three-fold increase in costs. This successful optimization of data management demonstrates the effectiveness of the platform.

Other benefits include an improved customer experience and increased operational efficiencies. Through the CRM campaign developed using the Data Intelligence Platform, Chorocmaeul has increased the number of active customers by at least 30% and improved the customer retention rate by 10%. Additionally, the inconvenience in service use has been considerably reduced through fast data management and sharing, further enhancing efficiency in service provision. For instance, Chorocmaeul can now swiftly take real-time business actions, such as notifying customers of unused coupons, resulting in a threefold increase in the customer conversion rate and more than doubling the return on investment.

After optimizing inventory management and improving demand prediction, Chorocmaeul has increased their handling rate to more than 80%. They expect to enhance safety margins and flexibility by further strengthening their streaming functions with Databricks Delta Live Tables, enabling real-time monitoring of sales status for safety stock management. Sooa Kwon highlighted the promising partnership between Jeongyookgak, Chorocmaeul and Databricks, “We plan to apply analytical model results and insights from the Databricks Platform to our operational database and leverage this information within our services.”