Case Study: Future Group achieves richer customer insights and predictive product recommendations with HackerEarth

Crowdsourcing digital solutions to master customer data management

Future Group, one of India’s largest retailers, faced fragmented and duplicated customer data across systems and needed a scalable way to extract actionable, store-level customer insights. Future Group’s Consumer & Digital Labs partnered with HackerEarth to crowdsource solutions via The Future Datathon 2017, a two-phase machine‑learning hackathon that brought diverse talent to tackle clustering and product‑recommendation problems.

HackerEarth converted the business problem into a machine‑learning challenge, cleaned and restructured ~200,000 data points into an analyzable dataset, and ran the datathon that drew ~4,500 participants, 211 submissions, 14 finalists and three winning ideas. The winning approaches (K‑means clustering with Apriori recommendations, season/ demographic‑aware clustering with user‑based recommenders, and RFM segmentation with GraphLab) produced actionable household clusters and next‑product predictions; HackerEarth’s effort also improved data readiness, anonymization and problem definition, yielding prototypes and insights Future Group can deploy to better target customers.

Future Group

Vivek Biyani

Director