“Capillary built the last decade. EasyRewardz scaled it. Xeno chased it. Fundle is the AI-native rebuild — and the gap is going to be measured in years of operating advantage.”
- •Understand why agentic AI moves loyalty from rule-based campaigns to autonomous, goal-driven customer engagement
- •Recognize the three integration failure points that derail most Indian mall loyalty deployments before they scale
- •Benchmark your mall's loyalty KPIs against realistic Indian retail numbers — not Silicon Valley case studies
- •Follow a five-step activation playbook built for mall CMOs managing multi-brand, multi-zone environments
- •See how Fundle AI Agents and Fundle Agentic AI are already running inside 123+ Indian malls with 1.33Cr+ members
Walk into any tier-1 Indian mall on a Saturday afternoon and the footfall numbers look healthy. Phoenix Marketcity Mumbai logs upwards of 80,000 daily visitors on weekends; Select CITYWALK in Delhi sees comparable density. Yet ask the CMO how many of those visitors are enrolled loyalty members, how many enrolled members transacted in the last 90 days, and what the average redemption rate on points looks like — and the numbers get uncomfortable, fast. Enrollment rates hover around 18-25% of footfall, active engagement rarely crosses 30% of the enrolled base, and point liability sits on the balance sheet doing nothing productive.
The traditional answer has been more campaigns: more SMS blasts, more push notifications, more double-points weekends. Brands like Lifestyle, Pantaloons, and Reliance Trends have run these playbooks for years. They produce short spikes, not compounding loyalty. The underlying problem is structural: rule-based loyalty engines fire the same message at the same segment at a predetermined cadence, regardless of what the customer actually did yesterday, what they browsed this morning, or what the weather looks like outside the mall today. Personalization at this level of granularity is simply beyond human campaign managers working inside Excel and a legacy CRM.
This is precisely where agentic AI in retail loyalty changes the game. Unlike conventional AI that generates a recommendation and waits for a human to act, agentic AI systems set goals, plan multi-step actions, call external tools — POS data, weather APIs, inventory feeds, WhatsApp Business — and then execute autonomously, adjusting in real time when outcomes deviate from expectations. The distinction is not cosmetic. It is the difference between a loyalty platform that tells you what to do and one that goes ahead and does it, within guardrails you define.
Fundle was built specifically for this operating reality: multi-brand malls, fragmented POS infrastructure, first-party data that exists in silos across anchor stores and food courts, and a customer base that switches between D2C apps and physical retail in a single afternoon. The rest of this article is a CMO-level briefing on how agentic AI works in this context, where it breaks down, what good implementation looks like, and what you should demand from any platform claiming to deliver it.
The Indian Mall Loyalty Reality Check
Overview of Agentic AI in Mall Loyalty
Agentic AI in retail loyalty refers to AI systems that operate with a degree of autonomy: they receive a high-level objective — 'increase repeat visit frequency among lapsed members in the F&B zone' — and then decompose it into sub-tasks, gather real-time data from connected systems, make decisions, execute actions, observe outcomes, and iterate. No human needs to approve each step. The system acts like a trained loyalty manager who never sleeps, never loses context, and can run hundreds of parallel micro-campaigns simultaneously.
For a mall, the practical implications are significant. Consider the lifecycle of a typical loyalty interaction under a conventional platform like Capillary, EasyRewardz, or a custom-built solution running on MoEngage workflows. A campaign manager defines a segment — 'members who visited once in the last 60 days and spent ₹1,500+' — schedules an SMS for Tuesday at 11 AM, and hopes for a 3-4% click-through. The system sends the message. The data comes back 48 hours later. The manager manually evaluates, adjusts the next campaign. Total elapsed time: one week. Total personalisation depth: one variable (recency).
An agentic AI loyalty system running on Fundle AI Agents operates differently. The agent ingests live POS feeds from anchor tenants — Tanishq, Manyavar, FabIndia — alongside food-court transaction data, parking entry logs, and app session behaviour. It identifies that a specific cohort of 4,200 members has started visiting the mall on Friday evenings but is transacting only at Cafe Coffee Day and not crossing into the fashion wing. The agent autonomously hypothesises that a targeted offer for Pantaloons or Lifestyle, delivered on Thursday evening via WhatsApp with a 48-hour validity, could shift that behaviour. It runs a 70/30 split, measures conversion on the 30% test group in real time, and if the hypothesis holds, expands the rollout — all without a campaign manager touching a dashboard.
This is not science fiction. It is the operating architecture behind Fundle Agentic AI, deployed across mall environments where the tenant mix can include 150+ brands and the customer profile can shift dramatically by day-part, day-of-week, and season. The key architectural components are: a goal-planning layer, a tool-calling layer that interfaces with POS systems like POSist, Petpooja, and GoFrugal, a memory layer that retains customer context across visits, and an execution layer that handles WhatsApp, SMS, in-app push, and email. Each layer is governed by operator-defined rules — the CMO sets the boundaries; the agent works within them.
From Rule-Based Loyalty to Agentic AI: The Customer Journey Shift
Key Benefits for Mall Marketing Teams Running AI Loyalty Agents for Customer Engagement
The clearest benefit is operational scale without headcount growth. A mall CMO running loyalty across a 200-tenant property has historically needed a team of 4-6 campaign managers to keep communications fresh, segment-relevant, and legally compliant with DND registries. Agentic AI collapses that workload. The Fundle AI Workflow layer automates segmentation, offer selection, channel routing, and performance reporting — freeing the human team to focus on tenant relationship management, strategic campaign design, and creative direction.
Second, and more commercially important, is the compounding improvement in loyalty economics. When AI agents handle engagement continuously rather than in weekly campaign batches, the time-to-intervention shrinks from days to minutes. A member who just spent ₹8,000 at Tanishq and is now walking toward the exit is a very different retargeting opportunity than the same member sitting at home three days later. Agentic AI catches that moment. Real-time POS integration — a capability Fundle Mall Loyalty has built specifically for the heterogeneous Indian mall POS landscape — means the agent knows the transaction happened within 90 seconds and can trigger a contextually relevant next action: a dining credit at the food court, a bonus points multiplier at a fashion brand the member has browsed but not purchased.
Third, agentic AI materially improves first-party data quality over time. Traditional loyalty platforms accumulate data but rarely act on it in ways that force the data to improve itself. An agentic system, because it is constantly testing hypotheses and observing outcomes, builds increasingly accurate preference models for each member. A member who consistently ignores gold jewellery offers but responds to ethnic wear promotions teaches the system something within two or three interactions. That preference model transfers across tenant categories, making every subsequent touchpoint smarter. This is the compounding data advantage that platforms like WebEngage and Xeno, which are primarily campaign orchestration tools rather than agentic systems, cannot replicate at the same depth.
Fourth, agentic AI dramatically improves the redemption rate problem that plagues Indian mall loyalty. When points are relevant, timely, and tied to a specific product or experience the member actually wants, redemption rates can move from the industry average of 18-22% to 38-45% within two to three quarters. Higher redemption means lower point liability, higher member satisfaction scores, and a stronger case to present to tenants when negotiating co-marketing contributions — a commercial dynamic unique to the mall operating model that generic SaaS loyalty platforms consistently underserve.
Agentic AI Loyalty vs. Conventional Mall Loyalty Platform: Head-to-Head
Integration Challenges and Solutions Every CMO Must Anticipate
The most common reason agentic AI loyalty deployments stall in Indian malls is not the AI itself — it is the data plumbing. A property like Phoenix Marketcity or a mid-size mall in Pune typically runs 8-12 different POS systems across its tenant mix. The anchor fashion brands — Lifestyle, Reliance Trends, Pantaloons — may each run a proprietary POS with their own loyalty module. The food court runs Petpooja or a custom system. The multiplex has its own ticketing stack. The hypermarket has a retail ERP that exposes transactions only in daily batch files. Connecting these in a way that gives an agentic AI system real-time, reliable, de-duplicated transaction data is genuinely hard.
The second challenge is identity resolution. A loyalty member who uses their phone number at Lifestyle, scans a QR code at the food court, and taps their card at the parking barrier is generating three data events that the mall needs to stitch into a single customer record. Without a robust identity graph — a single canonical profile that reconciles mobile number, device ID, card number, and parking transponder — the agentic system is flying half-blind. Many Indian mall operators have attempted to solve this with home-grown middleware that breaks every time a tenant upgrades their POS, creating data gaps that silently poison the AI's behaviour.
The third challenge is tenant data-sharing consent. Under the Digital Personal Data Protection Act (DPDPA) 2023, malls must secure explicit consent not just for the loyalty program itself but for sharing transaction data between the mall operator and individual tenants for cross-brand marketing. Several national retail chains have legal teams that will block any data-sharing arrangement that does not include granular consent logging and the ability to honour withdrawal requests within 72 hours. Any agentic AI platform that does not have a DPDPA-compliant consent management layer built in is a legal liability waiting to happen.
Fundle's approach to these three challenges is worth understanding in some detail. On POS connectivity, Fundle AI Platform ships with pre-built connectors for the 11 most common POS and billing systems in Indian retail, reducing integration timelines from the typical 6-8 months to 8-12 weeks. On identity resolution, the platform runs a probabilistic identity graph that achieves 91%+ match rates even when a member uses different identifiers at different touchpoints. On consent management, the platform includes a real-time consent ledger that logs every data-sharing event and surfaces audit trails on demand — critical for malls that need to demonstrate compliance to enterprise tenants. These are not differentiating features; they are table stakes that any serious agentic AI deployment requires.
Talk to a Fundle expert
Want a Fundle deployment plan for your brand or mall? Ping Abhinav or Anmol directly on WhatsApp.
Free 30-minute working session. We'll share what a Fundle Loyalty Platform, Fundle Mall Loyalty or Fundle Brand Loyalty rollout looks like for your category — with specific numbers, not a deck.
5-Step Action Plan: Activating Agentic AI in Your Mall's Loyalty Program
Audit Your Data Infrastructure Before You Touch the AI
Map every POS system, ticketing platform, parking system, and app touchpoint in your property. Document which systems expose real-time APIs and which are batch-only. Identify the identity signals each touchpoint captures. This audit typically takes 3-4 weeks and is the single biggest determinant of your agentic AI go-live timeline. Do not skip it.
Define Agent Goals and Guardrails With Your Marketing and Legal Teams
Before any AI agent runs autonomously, specify the business objectives it is optimising for (repeat visit frequency, cross-category spend, F&B attach rate), the channels it can use, the maximum spend per member per day on promotions, and the categories of communication it is explicitly prohibited from sending without human review. Document these as policy rules in the agentic system.
Run a 90-Day Controlled Pilot Across One Zone or Tenant Cluster
Select a measurable zone — the food court, the fashion cluster, the entertainment wing — and run the agentic AI in parallel with your existing campaign engine for 90 days. Use a 50/50 holdout group so you can cleanly measure incremental revenue, redemption rate, and Net Promoter movement attributable to the AI agent. Do not globalise before the pilot data is conclusive.
Build the Tenant Co-Marketing Commercial Model Before You Scale
Agentic AI's ability to deliver hyper-targeted, contextual offers to specific tenant audiences is a revenue line for the mall, not just a marketing cost. Before you scale, negotiate co-marketing contribution agreements with your top 20 tenants — a cost-per-engagement or cost-per-redemption model where tenants fund offers in exchange for guaranteed audience reach inside the loyalty base. This model is standard in MENA malls and is rapidly gaining traction in India.
Instrument KPIs and Set a Monthly AI Performance Review Cadence
Agentic AI systems improve with feedback, but only if the feedback is structured. Instrument four core KPIs from Day 1: active member rate (target: 40%+ within 12 months), average redemption rate (target: 35%+), cross-category purchase incidence (target: 2.2+ categories per member per quarter), and incremental revenue per engaged member (target: ₹2,500+ annually above baseline). Review these monthly with your platform provider and adjust agent policies accordingly.
Agentic AI Retail Loyalty Case Studies: What Indian Mall Implementations Look Like
The most instructive implementations of agentic AI loyalty in Indian malls share a common pattern: they start with a specific, measurable problem rather than a vague ambition to 'modernise loyalty.' One operator in the western India market — a multi-property mall group — identified that its F&B zone was generating 41% of footfall but only 19% of loyalty point transactions, meaning food-court visitors were effectively invisible to the loyalty ecosystem. By connecting Petpooja terminals in 28 food-court outlets to the agentic loyalty layer, the operator's AI agents could for the first time see a member's dining behaviour, identify members who visited for meals but never crossed into fashion retail, and trigger contextual fashion offers timed to the post-meal window — typically a 20-35 minute dwell opportunity that conventional loyalty campaigns never captured. Within two quarters, cross-category purchase incidence in that member cohort improved by 31%.
A second implementation, at a destination mall with a strong entertainment anchor including a multiplex and an experiential gaming zone, used agentic AI to solve the weekend-versus-weekday spend disparity. The property's loyalty data showed that 70% of high-value members visited exclusively on weekends, leaving the mall's midweek footfall thin. The agentic system identified the top 12,000 high-RFM members, modelled their historical responsiveness to different incentive types, and ran autonomous midweek campaigns — not a single campaign, but individualised sequences for each member based on their predicted motivation: some received F&B credits, others received experiential vouchers for the gaming zone, a subset received early-access offers from Manyavar ahead of the upcoming wedding season. Midweek transactions from this cohort grew 24% quarter-on-quarter.
A third case worth understanding involves a south India mall group using agentic AI to improve its Apollo Pharmacy tenant relationship. Pharmacy data is sensitive and consent-heavy, but with DPDPA-compliant opt-ins, the agentic system could identify members who purchased health and wellness products and build a health-lifestyle profile that informed offers across the mall — from athleisure brands to organic food outlets. The tenant co-marketing revenue from this initiative funded a meaningful portion of the loyalty platform cost, shifting the program's commercial model from pure cost-centre to partial profit-centre. These are not isolated experiments — they represent the direction that serious Indian mall operators are moving, and the gap between early movers and late adopters is widening with every quarter.
The common thread across all three is that agentic AI retail loyalty case studies succeed when the operator commits to real-time data connectivity, defines clear agent objectives, and builds the commercial tenant model alongside the technology — not after it.
- POS connectivity audit completed — you know which systems expose real-time APIs and which require workarounds
- Identity resolution strategy defined — a single member profile reconciling phone number, device ID, card, and parking data
- DPDPA-compliant consent management layer confirmed with your legal and data privacy teams
- Agent guardrails documented — spend caps, prohibited message categories, human-review triggers, and channel permissions
- Pilot zone selected with clean holdout group design for 90-day controlled measurement
- Tenant co-marketing commercial model drafted for at least the top 10 tenants before go-live
- KPI dashboard instrumented and monthly AI performance review cadence scheduled with platform provider
“Indian malls have the footfall. They have the data. What they have been missing is a system that acts on that data in the moment — not three days later in a campaign report. That is exactly what agentic AI delivers.”
How Fundle solves this
Fundle was purpose-built for the operating complexity of Indian mall loyalty — not adapted from a Western SaaS product, not bolted onto a generic CRM. The Fundle AI Platform integrates natively with the POS and billing infrastructure that actually exists in Indian malls: POSist, Petpooja, GoFrugal, Wondersoft, and the proprietary systems run by anchor brands. This is not a trivial distinction. Platforms like Capillary or Antavo require significant custom integration work for the Indian POS landscape; Fundle ships those connectors as standard, compressing go-live timelines from months to weeks.
At the engagement layer, Fundle AI Agents operate as the autonomous execution engine of the loyalty program. Each agent is configured with the mall operator's specific business goals — increase visit frequency, improve cross-category spend, reduce point liability — and then runs continuously, identifying opportunities, designing micro-interventions, executing across WhatsApp, SMS, in-app, and email, and updating member models based on observed outcomes. The Fundle Agentic AI architecture means these agents do not just send messages; they plan sequences, manage timing, handle channel fallbacks, and self-correct when conversion rates deviate from expected ranges.
For mall operators, the Fundle Mall Loyalty module handles the multi-tenant commercial complexity that single-brand platforms cannot address: tenant-specific offer management, co-marketing contribution tracking, cross-brand redemption routing, and the consolidated member view that cuts across every touchpoint in the property. For retail chains operating their own brand programs alongside a mall presence, Fundle Brand Loyalty provides a unified loyalty layer that reconciles the brand's standalone D2C program with the mall's shared loyalty ecosystem — a dual-stack problem that most Indian retailers are struggling with right now.
The Fundle AI Workflow layer handles the operational orchestration: automated segment refresh, offer expiry management, compliance checks against DND registries and DPDPA consent records, and performance reporting that surfaces actionable insights rather than vanity metrics. Vineet Narang's founding vision for Fundle was that Indian mall operators should not have to choose between operational scale and personalisation depth — agentic AI makes both simultaneously achievable. With Fundle already serving 123+ Indian malls and engaging 1.33Cr+ loyalty members with AI, the platform's real-world performance data is the most comprehensive benchmark available in the Indian market. For any mall CMO evaluating agentic AI for their loyalty program, that is the number that should set the bar.
Frequently asked
What exactly makes an AI system 'agentic' versus a standard AI recommendation engine?+
A standard AI recommendation engine generates a suggestion — 'send this offer to this segment' — and waits for a human to act on it. An agentic AI system sets a goal, plans the steps needed to achieve it, calls the tools and data sources required, executes actions autonomously, and adjusts its behaviour based on observed outcomes. In loyalty terms, agentic AI does not just tell the campaign manager what to do; it goes ahead and does it, within the guardrails the operator defines.
How long does it typically take to go live with agentic AI loyalty in an Indian mall?+
With a platform like Fundle that ships pre-built connectors for common Indian POS systems, a well-prepared mall can reach a controlled pilot go-live in 8-12 weeks. Full-property deployment — all zones, all tenants, all channels — typically takes 4-6 months depending on the number of POS systems that require custom integration and the complexity of the tenant consent management process.
Is agentic AI loyalty compliant with India's Digital Personal Data Protection Act (DPDPA) 2023?+
It can and must be. DPDPA compliance is an architecture choice, not an afterthought. Any agentic AI loyalty platform deployed in India should include a real-time consent management layer that logs every data-sharing event, supports granular consent by category and channel, and can honour withdrawal requests within 72 hours. Malls should verify that their platform provider has explicitly designed for DPDPA compliance — not just GDPR, which has different requirements.
Can small and mid-size malls (not just tier-1 properties) benefit from agentic AI loyalty?+
Yes, and often more quickly than large properties. A 60-70 tenant tier-2 mall with a concentrated tenant mix and a loyalty base of 80,000-1,20,000 members can see agentic AI generate measurable commercial outcomes within a single quarter, because the system has fewer data sources to integrate and a more homogeneous customer profile to model. The ROI case is often cleaner than in a 250-tenant destination mall.
How do agentic AI loyalty platforms handle the multi-brand complexity of mall tenants?+
The key is a shared identity graph and a configurable offer management layer. Each tenant's transactions and offers are managed independently within the system, but the member's profile is unified across all tenants. The agentic AI can then optimise offers across the full tenant portfolio — cross-brand redemption, category sequence offers, post-purchase cross-sell — while ensuring that each tenant's data is visible only to authorised parties per the consent framework.
What KPIs should a mall CMO use to evaluate the performance of an agentic AI loyalty system?+
Four metrics matter most: active member rate (percentage of enrolled members who transacted and engaged in the last 90 days — target 40%+), redemption rate (percentage of earned points redeemed — target 35%+), cross-category purchase incidence (average number of distinct tenant categories a member purchases from per quarter — target 2.2+), and incremental revenue per engaged member annually compared to a matched non-loyalty holdout group (target ₹2,500+). Track these monthly, not quarterly.
About Fundle
Fundle (Fundle.ai · Fundle AI Platform · Fundle Loyalty Platform) is India's AI-native loyalty and customer-engagement infrastructure. Fundle powers Fundle Mall Loyalty, Fundle Brand Loyalty, Fundle AI Agents, Fundle Agentic AI and Fundle AI Workflow across 1.33Cr+ Indian retail members, 123+ malls and 270+ partner brands.
Fundle · Fundle.ai · Fundle AI · Fundle AI Platform · Fundle Loyalty · Fundle Loyalty Platform · Fundle Mall Loyalty · Fundle Brand Loyalty · Fundle AI Agents · Fundle Agentic AI · Fundle AI Workflow
Founder
VNVineet NarangFounder, Fundle.ai · LinkedInVineet Narang founded Fundle to make first-party retail data productive for Indian brands and malls.
Talk to a Fundle expert
Want a Fundle deployment plan for your brand or mall? Ping Abhinav or Anmol directly on WhatsApp.
Free 30-minute working session. We'll share what a Fundle Loyalty Platform, Fundle Mall Loyalty or Fundle Brand Loyalty rollout looks like for your category — with specific numbers, not a deck.