Project 03 · Retail Expansion Intel · demo

Where in India should a retailer open its next store?

During my KPMG internship, a retail client wanted to know where to expand. The honest answer was sitting in public data: every competitor's store locator, plus the census. It just needed putting on one map.

Focus: The data

Cities as dots · illustrative layout

Step 1 of 6

Start with an empty map.

Every dot is a city. The client's question was simple: which of these should get the next store?

Step 2 of 6

Add every competitor's store.

I scraped the public store locators of 8 chains, cleaned the addresses & geocoded each one. 2,600+ dots later, India looks very different.

Step 3 of 6

Now count them.

Rolled up by city & district, the big metros are crowded. Nobody's surprised by that part.

Step 4 of 6

Layer in the neighbourhood.

Stores don't open in a vacuum, so we pulled in secondary data on top: census population, petrol pumps, malls, even Domino's outlets as a signal of local spending.

Step 5 of 6

Score every spot.

We combined all of it into a scoring system that grades each district as a place to open the next store. Lots of people & spending, few competitors: those light up.

Step 6 of 6

Out comes a shortlist.

Each district gets an opportunity score, so the conversation shifts from 'where do we feel good about' to a ranked list with reasons.

Stores mapped: 2,600+
Chains: 8
Output: Opportunity score by district
Data: Public store locators + census

Built with Python · Streamlit · Geocoding · Census data

Next project

Group Seat Finder →

Four friends, four budgets, one row. Finds the theatre seat block that makes the whole group happiest.

Setting the type…

The Yash Hour