ROI Model
Win-back campaign ROI for the At-Risk segment — three recovery scenarios, sensitivity analysis, and Indian retail network extrapolation.
At-Risk Segment Baseline
| Metric | Value (£) | Value (₹) | Source |
|---|---|---|---|
| At-Risk customers | 661 | — | Observed — RFM segmentation |
| At-Risk segment historical revenue | £88,223 | ₹94.4 lakh | Observed — dataset |
| Average revenue per At-Risk customer | £133 | ₹14,231 | Calculated (£88,223 ÷ 661) |
| Forward revenue at risk (proxy) | £88,223 | ₹94.4 lakh | Modelled — assumes historical run rate if unrecovered |
£1 = ₹107. Revenue figures reflect historical spend across the dataset period (Dec 2010–Dec 2011), not guaranteed forward revenue.
Campaign Cost Assumptions
Indian market benchmarks (deployment context)
| Cost Item | Assumption | Basis |
|---|---|---|
| Email platform cost per send | ₹2–₹5 per email | Mailmodo, Netcore, SendGrid India pricing (1,000+ volume) |
| Creative and copywriting | ₹15,000–₹30,000 (one-time) | 3-email drip sequence, mid-market agency rate India |
| Analytics and reporting | ₹10,000 (one-time) | 2 analyst days at ₹5,000/day |
| Total campaign cost (661 sends × ₹5 + fixed) | ₹58,305 | Modelled estimate — ~£545 GBP equivalent |
Three Win-Back Scenarios
ROI formula: (Revenue Recovered − Campaign Cost) / Campaign Cost × 100. Recovery rate = % of At-Risk customers placing at least one qualifying order within 90 days. Benchmarks sourced from Klaviyo and Salesforce Marketing Cloud consumer retail research (2022–2024).
Sensitivity Analysis
Net revenue recovered (₹) across varying recovery rates and campaign costs
| Recovery Rate | Campaign ₹40K | Campaign ₹58K (Base) | Campaign ₹1L |
|---|---|---|---|
| 10% | ₹9,03,954 | ₹8,85,649 | ₹8,43,954 |
| 15% | ₹13,75,931 | ₹13,57,626 | ₹13,15,931 |
| 20% | ₹18,47,815 | ₹18,29,510 | ₹17,87,815 |
| 25% | ₹23,15,557 | ₹22,97,252 | ₹22,55,557 |
Break-even recovery rate at ₹58K campaign cost ≈ 0.7% (5 of 661 customers). The campaign is cash-positive above this threshold.
Network Extrapolation — Indian Retail
Applying the same methodology at scale across two Indian retail reference networks. All figures are modelled estimates — explicitly labelled.
DMart — ~350 stores
Avenue Supermarts Ltd (2024, publicly reported store count)
Reliance Retail — ~18,000 stores
Smart Bazaar, Trends, Jio Points, Reliance Digital (2024)
DMart does not operate a traditional loyalty programme as of 2024 — this extrapolation assumes a DMart-equivalent retailer with loyalty enrolment. Store counts from publicly reported 2024 annual reports. Customer-per-store and revenue-per-customer figures are modelled estimates (confidence: low). See Sources for full disclosure.
India Compliance Note — DPDP Act 2023
The Digital Personal Data Protection Act 2023 governs how Indian businesses collect, store, and use customer data for marketing purposes. A win-back campaign of this type would require explicit consent from At-Risk customers before sending email communications, and a clear opt-out mechanism. The agentic workflow architecture in this project includes a human approval gate — a natural integration point for a compliance check before any campaign is dispatched.