Indeed Reduces Costs by 27% in 6 Months with MongoDB

Indeed helps millions of people search and apply for new jobs online, and its platform is built on more than 500 TB of data. When the company saw an opportunity to improve its service by migrating to the cloud, it moved from its self-managed solution to Amazon Web Services (AWS) and MongoDB Atlas, MongoDB’s fully managed, multicloud developer data platform.

After completing the migration in November 2023, Indeed discovered that its costs had increased significantly. It subsequently set a goal to reduce costs by 19% and began using MongoDB Atlas features and support to achieve that goal.

The company worked with the MongoDB Atlas Technical Services team to analyze its 600 clusters. Indeed found that 85% of its MongoDB Atlas costs came from specific clusters and began its cost optimization strategy by addressing those clusters.

“We started with the cluster that was burning the most money, and then we went in phases, working with the product teams to optimize it and track costs,” said Manan Shah, cloud data engineer at Indeed. “We also worked with MongoDB Atlas Technical Services, who helped build a tool that analyzes a cluster and provides recommendations.” The technical services team recommended using lower tiers of compute instances for several clusters, resulting in immediate cost savings.

Manan Shah, Cloud Data Engineer

Indeed optimized data storage by increasing the storage tier and switching from provisioned input/output operations per second (IOPS) to standard storage volumes. “We scaled IOPS through storage, and that was more cost-effective than going with the provisioned IOPS,” said Chitradeep Chakraborty, senior manager of cloud database and messaging engineering at Indeed.

In one example, the company’s storage spending shrunk by more than 30%. In addition, Indeed reduced the number of replicas in its multiregional clusters and removed idle resources. With these further optimizations, it eliminated unnecessary spending and reduced total spend by an additional 3%.

In just six months, the company reduced its total costs by 27%—far exceeding initial goals. Subsequently, Indeed moved away from third-party backup storage and adopted the cross-region data transfer feature of MongoDB Atlas. It also used the MongoDB Atlas’s network compression capabilities to compress data transfers and further reduce costs. Without giving up any of its performance, storage capacity, or support benefits, the company streamlined infrastructure efficiency.

Indeed plans to continue making improvements to other aspects of its infrastructure, such as data backups. Overall, the company is spending less while delivering the same great experience to users. “We switched from our custom solutions to doing it all within MongoDB Atlas,” said Shah. “A significant amount of our costs was saved because of this combination of solutions.”

Manan Shah, Cloud Data Engineer