Kmong achieves 30% higher conversion with AWS


The Challenge

When Kmong began, small business design-related transactions dominated, but as it shifted to larger companies its service categories expanded into marketing, IT, and business consulting. The number of registered freelancers and service categories grew too, and transaction volumes went rapidly.

To expand its services and accommodate growth, Kmong grew from 25 staff to 81, and from four teams to 10. As the teams became more specialized, they wanted to refer to data for important decisions. Kmong then created a dedicated data team with a data analyst, a server engineer, and a deep-learning expert. The data team supports data analysis, including business planning, promotions, and development of big data, and artificial intelligence (AI)-based services such as customized or personalized recommendations.

“We found that fulfilling all these responsibilities was a challenge for our three-person data team, so we turned to the cloud,” says Park Jae-young, Kmong CTO.

"Thanks to our funnel improvement, we gained deeper customer insight that led to a 30% increase in purchasing conversion and a 40% decrease in churn."

Park Jae-young, CTO, Kmong

  • About Kmong
  • Founded in 2012 to connect buyers and sellers of freelance services, Kmong has grown into Korea’s first business services marketplace, with 170,000 registered experts in 11 categories. Kmong reports more than one million transactions as of 2019, with 98% customer satisfaction and strong loyalty.

  • Benefits
    • 30% higher conversions
    • 40% drop in customer churn
    • Data-driven decision culture
    • Increased business agility
    • Higher customer satisfaction
  • AWS Services Used

Why Amazon Web Services (AWS)

Kmong’s data team wanted to build a data pipeline to improve the efficiency of data collection, processing, analysis and machine learning. This helped it focus on developing important new services like AI and data science rather than expending resources to process repetitive tasks.

“We compared various vendors, including Google Cloud Platform, and concluded that Amazon Web Services (AWS) was far superior in terms of operation and cost efficiency. It simply provides the best option for linking various log data, product-related data, and statistical data," says Mr Park.

In 2018, Kmong launched its data pipeline based on Amazon Aurora, Amazon EMR, Amazon Redshift, Amazon QuickSight, and Amazon Simple Storage Services (S3). The AWS platform allows Kmong to easily process big data analysis and make better business decisions. Kmong also integrated Microsoft Power BI and Tableau software with QuickSight to provide an intuitive end-user analysis tool.

“With the expertise of Amazon Professional Services, we completed the implementation in only three months,” Mr Park says. “The AWS solution was well-directed, so there was no obstacle to building our data pipeline.” Kmong also plans to leverage the platform to support the DevOps project while switching the platform base to microservice architecture.

The Benefits

Data-centric decision-making is a huge help in Kmong’s business growth. Before the implementation of the data pipeline, Kmong held 1TB of data. By 2019, it handled 20TB of data in real time with the AWS environment. Various teams can now make independent, data-oriented decisions by pulling, analysing, and visualizing data themselves.

"Thanks to our funnel improvement, we gained deeper customer insight that led to a 30% increase in purchasing conversion and a 40% decrease in churn," Mr Park explains.

In addition, the analysis platform allowed the data team to focus on strengthening AI capabilities like customized recommendation, benefitting Kmong’s performance. “Big data is used as a data set to run machine learning, so customers can get sophisticated, personalized services,” says Mr Park. “The purchase conversion rate for users who viewed our content recommendations is now 400% higher than those who viewed our normal content. This allows us to quickly launch a new service that matches buyers and sellers’ preferences or purchasing behaviours, and I hope this will push the rate even higher."

Kmong expects the establishment of a DevOps-based analysis platform will further enhance its agility and competitiveness. “Our service categories will become more diverse as our business grows. To keep pace with our growth, we’ll run DevOps in the AWS environment to speed up service expansion." Mr Park says.

Learn More

To learn more, visit Data Lakes and Analytics on AWS