Milkbasket increases team efficiency by 40% with APM and alerts

As India’s first and largest daily micro-delivery service, Milkbasket was built on the unique Indian habit of getting fresh milk delivered at home every morning. Founded in 2015, the service now fulfills the grocery needs of an entire household every day before 7am. To enable frequent and frictionless buying, Milkbasket has innovated flexi-ordering with a zero checkout and contactless delivery model - both firsts for India’s e-commerce industry.

Challenge: Meeting the demands of a growing customer base

E-commerce players are under phenomenal pressure to deliver products swiftly and smoothly in spite of a swathe of challenging conditions such as strained supply chains.

A New Relic customer since inception, Milkbasket uses the observability platform to deliver a best of breed product to its customers, looking at insights into basic health checks, and highlighting features in use by the end-user.

“Milkbasket uses New Relic to gather insights into consumer team performance, which has helped us have strong response times,” explains Devika Razdan, Head of Product & Technology, Milkbasket.

“Because we require end-to-end observability, we chose to focus on New Relic as our centralised platform, and expanded our adoption to include APM and mobile as a result.”

Solution: Enhanced observability for scale

During the height of the COVID-19 pandemic, the Milkbasket team was able to handle a spike in demand that was almost five times its usual traffic–without breaching its Service Level Agreements (SLAs) of less than 100 milliseconds. As a result, around 150,000 new users were able to join the platform.

More recently, Milkbasket has successfully met the demand generated by its savings days sales events; seeing a two to three time increase in the number of orders it receives. By providing support in issue identification and root cause analysis, New Relic has supported Milkbasket to handle this growth and associated spikes in demand.

"In one instance, a few of our customers told us that the search page was loading slowly. We checked the peak hour load and the throughput–it was all looking good. But through New Relic, we were able to carry out stack trace analysis of p99 and p95 requests and identify issues around caching in a specific scenario. With the root cause of the problem identified, we could immediately fix it," Devika says.

New Relic has also enabled Milkbasket to adopt proactive measures. For example, when traffic increases, Milkbasket can scale its offerings to meet demand.

Devika’s team holds a bi-weekly meeting where a report or dashboard, based on data from New Relic, is prepared that includes a snapshot of usage. If an issue has created a surge, Milkbasket can quickly drill down and find the source of the problem.

Today, according to Devika, Milkbasket “relies on New Relic so much” that whenever the team needs to track a new release or production update, it will use New Relic to ensure that its tech stack is running smoothly and health metrics are controlled.

"We have a performance testing team carrying out load testing using (Apache) JMeter. Through New Relic, they figure out the trend: what endpoints have what kind of RPM associated. Based on that, they prepare a JMeter script and then test it using different loads. This helps us prepare for peak times.”

"New Relic is a real lifesaver. It's not just for day-to-day activities, it's for managing special events like marketing promotions. It’s a platform that we can rely on, which tells us upfront what issue has occurred or is about to occur," Devika says.

Impact: Improved SLAs, excellent customer experience

Measurable results from New Relic have been impressive. For example, Milkbasket has seen improved SLAs on consumer and operational experiences when searching for products or placing orders.

It has also seen improved success rates from multiple payment partners, with Milkbasket able to track both internal and third-party APIs and positively impact relationships with these partners.

Resolution findings and root cause analysis have greatly improved, with efficiency increasing by 40 percent and reduced MTTR. There’s less downtime and a highly available environment SLA of 99.99%.

Milkbasket has expanded its business and scaled with confidence; doubling the locations that it serves and expanding from 5 to over 23 cities during 2022 alone. In terms of plans for the future, the business hopes to offer its products to even more tier two cities in India, which will require more monitoring and alerting.

From a load perspective, Milkbasket is anticipating the spike in demand for its products to continue, and is confident they will be able to meet this, in partnership with New Relic.