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.
TBDBuild Duration
XXXXIndicative ROI
XXXXBuild Cost
XXXXValue
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.”