Swiggy Uses 350+ Features and Multi-Task MLP to Predict Customer Lifetime Value

Leela Kumili — Wed, 02 Sep 2026 13:55:00 GMT

Swiggy developed an in-house predicted lifetime value model using more than 350 pre order features and a multi task MLP for Food and Instamart. Adding order count as an auxiliary task reduced model parameters by 63% while improving predictive performance. The pLTV signal is used with Google Target ROAS bidding to optimize customer acquisition.

By Leela Kumili

InfoQ - Online Learning

Swiggy Uses 350+ Features and Multi-Task MLP to Predict Customer Lifetime Value