Swiggy is India’s largest food ordering & delivery platform. Founded in 2014, Swiggy has expanded its operations to over 500 cities1. In 2020, Swiggy launched initiatives like Instamart, the on-demand grocery delivery service & Genie, the instant package delivery service, which has been success stories of their own. Swiggy has recently forayed into newer avenues like Meat & Alcohol, with delivering convenience on-demand to the users at the centre of all the initiatives. .

Ishu Jain, Director of Analytics at Swiggy was interviewed for this case study. She leads the analytics initiatives across three key stakeholders, Customers, Delivery Partners & Restaurants. The Data Analytics team in Swiggy is over 100 members strong, working on ingesting data from various sources, transforming, and drawing insights from them.Using Snowflake as their data warehouse, the data team builds various statistical models & reports that empower different teams at Swiggy, to perform diverse sets of business tasks.

On the customer side, the team works on models for Revenue & Growth, Pricing,Promotions & Loyalty, Order Forecasting, discounts, marketing teams. The restaurant facing part of the team works with various restaurants, cloud kitchens, national &international chains. The delivery partner-facing team members work with route plans, sanity checks, governance etc. Almost 75% of the analysts time is occupied attending to data requests from various business & product teams, generating reports, doing RootCause Analysis. Another 15% is spent on building statistical models for different teams and the rest building dashboards.

Being data-driven in its approach, all the decisions at the organisation is backed by data. Each business vertical consumes data, reports & dashboards, for which they are dependent on the analytics team for their data needs. Thus, there was a huge dependency on the data analytics team, which would receive multiple requests from various departments. Limited by the time resource, the analytics team could only work on a limited set of these requests based on the priority. Leaving some teams, underserved or unserved in terms of data. Also, analysts faced issues of repeatable requests, where they found themselves working on the same set of reports &dashboards multiple times.

The team was on the lookout for a solution that solves their challenges. Their internal teams also worked on ways to automate repetitive tasks. A self-serve analytics solution was sought that would help them democratize data access to the users with little or no knowledge about database query languages.

Sprinkle pitched in the No-Code Analytics solution. A proof of concept was done with the team. Data models were built on Sprinkle for 5-6 priority areas for the driver side of the business. Once the models were built, they could be used to seamlessly generate reports on Sprinkle. The potential of Sprinkle’s Analytics solution was realised. In the next step, the solution was introduced to other verticals as well. Data models were then also built for other business verticals like Loyalty, Customer Experience, DriverExperience, application team on Sprinkle. The models built on Sprinkle were used to generate Segment reports & visualisations by analysts, business team, product team & other users.

Presently over 250 reports are generated each month using the Sprinkle Platform by the teams in Swiggy. Over 60 analysts regularly use Sprinkle to do analyses, build models& generate reports. Teams that were under-served or unserved can now themselves pull relevant data from the warehouse and generate insights. A mix of people who didn’t have access to data, or didn’t have time, use Sprinkle’s No-Code solution to generate required data, visualisations & draw insights. Thus, helping them to make data-backed business decisions.

The users are performing small analyses & generating required data themselves usingSprinkle. It frees up the time resources for the data analytics team, helping them concentrate on building advanced models & reports. The number of simple, repeatable requests have come down for the analytics team.