Calculating the Impact of an Engineering Project

Wanted Lab's Data Team calculates the value of its work as impact measured in KRW. Since it's hard for developers to express achievements as numbers, the team calculated impact based on how much resource a project reduced and what problem it solved.

The key case is the Weaver project. By introducing real-time data streaming infrastructure based on Apache Kafka, the team unified scattered event collection into a single endpoint and converted CDC, which had only been handled in batches, into real-time events.

Impact calculation was carried out in three ways.

• Development time savings: The team assumed that event SDK work accounts for 10% of overall work, and that Weaver would cut this roughly in half. Based on 8 squads and 3 developers, the impact of switching to a single event endpoint was calculated at ₩241,920,000 per year.
• Feature replacement cost comparison: The CDC real-time event conversion was calculated based on 5 squads, 2 developers per team, and 2 sprints per year, putting its value at ₩80,000,000 per year.
• Comparison with an external SaaS: Comparing against Company A, which builds a similar service, the team factored in 19 backend engineers, 7 data engineers, and an annual platform cost of ₩25,000,000. Compared with the cost of Wanted's 4-person data team operating this internally, the impact was estimated at roughly ₩4,460,200,000.
• Infrastructure usage savings: By switching from running constantly on ECS Fargate to a batch approach that runs for only 1 minute every hour, the team saved $8,269.44 per year, about ₩10,750,272, based on 32 consumers and an average of 3 partitions.

Ultimately, the key to calculating impact lies in accurately identifying the resources invested and the scope of impact resolved. Rather than getting the numbers perfectly right, the habit of consistently recording work hours and the effects of solving problems is the starting point for establishing calculation standards going forward.