UC-282
New-Branch Siting & Expansion Optimiser
Predicts ramp curves and cannibalisation for candidate branch locations and ranks where to deploy expansion capital for best return.
6-8Build Duration
TBDIndicative ROI
The Challenge
500-550 new branches are planned, each expected to carry ~INR 15.9 cr AUM, but siting decisions lack a modelled ramp or cannibalisation estimate against the existing 5,000+ network.
How It Works
- Ingests existing branch ramp histories and geo-demographics.
- ML predicts ramp and cannibalisation; return models rank candidates.
- Returns a ranked deployment plan with expected return per branch.
What It Removes
- Gut-feel branch siting
- Unmodelled cannibalisation
- Capital spread thinly
Input Data RequirementsExisting branch ramp curves, geo-demographics, branch economics, cannibalisation data
Output FormatRanked candidate sites, forecast ramp per site, cannibalisation estimate, capital deployment plan
“Every one of the 500-550 new branches opens where the model says the money is.”