Fundle.ai: AI Loyalty Agents Platform Transforming Indian Shopping Malls

Challenges in Mall Customer Engagement That Legacy CRM Cannot Fix

The challenges facing a mall marketing director in 2025 are not new in nature but have grown exponentially in complexity. The first and most acute is identity fragmentation. A shopper who buys ethnic wear at Manyavar, gets coffee at Cafe Coffee Day, and picks up spectacles at Lenskart in the same visit is generating three separate transaction events across three separate brand CRMs — none of which talk to the mall operator's system in real time. The mall's own loyalty app, if it exists, captures only those transactions where the customer remembers to scan. Attribution is broken before personalisation can even begin.

The second challenge is the unit economics of manual campaign management. A mid-sized mall with 180 tenants, four F&B zones, a multiplex, and a hypermarket has hundreds of potential micro-segments. Manually building campaigns for each segment — lapsed footfall, high-frequency low-spend, first-visit conversions, anchor-store loyalists drifting to competitors — requires a CRM team that most mall operators simply do not have the budget or talent to staff. Platforms like Capillary, EasyRewardz, and Xeno offer segmentation tools, but they still rely on human campaign managers to design the logic. That is a throughput bottleneck masquerading as a technology solution.

Third is the channel incoherence problem. Indian mall visitors interact across WhatsApp, SMS, push notifications, email, and increasingly through QR-triggered moments in-store. Each channel has different open rates, different optimal send times, and different content formats that resonate. A rule-based system applies a uniform channel hierarchy regardless of individual customer behaviour. The result is message fatigue — and Indian consumers uninstall apps faster than almost any other market; average retail app retention at Day 30 is below 14% according to AppsFlyer's India data.

Fourth, and perhaps most underappreciated, is the tenant-operator misalignment problem. Brands like Tanishq or FabIndia have their own loyalty programmes with significant member bases. They are reluctant to cede customer data to the mall operator. The mall operator, in turn, cannot personalise the cross-mall experience without that data. This creates a cold war of data silos that ultimately hurts both parties and hands the advantage to pure-play e-commerce. An AI loyalty agents platform must function as a trusted neutral orchestration layer — one that gives brands their own analytics and control while enabling the mall to build an aggregated engagement picture.

Role of AI Loyalty Agents in Personalisation and Automation

The term 'AI loyalty agents platform' is used loosely in the market, so it is worth being precise about what agentic AI actually does differently from a sophisticated rules engine. A rules engine is deterministic: if the customer has not visited in 45 days, send a win-back offer. An AI agent is goal-directed: given the objective of maximising this customer's next-30-day spend, autonomously select the timing, channel, offer type, creative variant, and redemption mechanism that maximises expected value — and then learn from the outcome to recalibrate the next action. The difference in outcomes is not marginal. It is structural.

In practice, agentic AI for retail loyalty operates across three interconnected layers. The first is the perception layer: ingesting signals from POS integrations (POSist, Petpooja, GoFrugal, Wondersoft), Wi-Fi dwell-time sensors, app behaviour, and transactional history to build a continuously updated customer model. The second is the reasoning layer: a large language model or a fine-tuned recommendation model that interprets those signals against business objectives — visit frequency targets, basket size goals, tenant category quotas — and proposes actions. The third is the execution layer: autonomous dispatch of communications across WhatsApp Business API, push, SMS, and in-app, with real-time offer fulfilment and redemption tracking.

Personalisation at this level unlocks outcomes that segment-level marketing cannot. Consider a customer profile at a Select CITYWALK: a 34-year-old woman, high-income, frequents the fashion floor but has never visited the wellness or fine-dining zones. A rule-based system would either ignore her category whitespace or send a generic 'Explore more' message. An AI agent recognises that her dwell-time data shows she pauses near the wellness corridor every visit but never enters, and that customers with similar profiles respond 3.1x better to an experiential trigger — a complimentary consultation offer from a wellness brand — than to a discount. The agent fires that specific trigger, tracks the conversion, and updates the model.

Retail loyalty automation with AI agents also dramatically changes the economics of tenant engagement. Instead of the mall marketing team manually building co-branded campaigns with individual tenants, the AI system can autonomously generate, A/B test, and optimise hundreds of tenant-specific micro-campaigns simultaneously. Brands like Lifestyle or Pantaloons, which run their own seasonal promotions, can push their offer parameters into the platform and the AI agents handle the targeting, timing, and channel selection — reducing campaign turnaround from two weeks to under four hours. This is the compounding productivity gain that makes an AI loyalty agents platform a strategic asset, not just a marketing tool.

Retail-Media Monetisation: How AI Agents Unlock Mall Advertising Revenue

There is a second-order opportunity in AI-powered mall loyalty that most operators have not yet formalised into a revenue line: retail media. When a mall operator builds a first-party audience of 500,000 opted-in, transaction-verified shoppers with rich behavioural profiles, that audience has significant advertising value to both tenant brands and external advertisers. Amazon's retail media business crossed $47 billion globally in 2024 — and the fundamental asset underneath it is precisely what mall operators are beginning to accumulate: a captive, purchase-intent audience with verified spend behaviour.

In the Indian context, the retail media opportunity inside malls is largely untapped. Brands like Tanishq, Lenskart, or Manyavar are already spending significant digital budgets on Meta and Google to reach audiences they could reach more efficiently inside a mall's first-party ecosystem — because those mall visitors are already within the conversion funnel, physically present or recently active. The challenge has been that mall operators lacked the audience infrastructure, the measurement layer, and the campaign tooling to make a credible retail media pitch. An AI loyalty agents platform provides all three.

The Fundle AI Platform includes a retail-media module — Fundle Reach — that enables mall operators to package their first-party audiences into targetable segments and sell sponsored placements to tenant brands and external advertisers via a self-serve or managed interface. A jewellery brand like Tanishq can bid to reach female shoppers aged 28–45 who visited a competitor jewellery store in the last 60 days and have a verified household income above ₹15 lakh. That targeting precision is impossible on Meta without significant data leakage and model degradation from privacy changes. Inside Fundle's walled garden, it is native.

For mall operators, this transforms the loyalty programme from a cost centre into a profit centre. A mall with 600,000 active loyalty members generating 400 million data points per month can realistically expect a retail media revenue run-rate of ₹8–18 crore annually once the Fundle Reach infrastructure is properly commercialised. That is not a rounding error — for many Tier-2 mall operators, that figure represents 12–20% of total non-rental income. It also fundamentally changes the internal conversation about loyalty programme ROI: the question is no longer just 'what is the incremental visit frequency?' but 'what is the CPM yield on our audience?' This is the transformation that Fundle Mall Loyalty is designed to enable.

Metrics to Track Loyalty Engagement Success in AI-Powered Mall Programmes

One of the most common mistakes mall marketing directors make when evaluating loyalty programmes is measuring the wrong things. App downloads, total members enrolled, and points issued are vanity metrics — they correlate poorly with actual business outcomes. The six metrics below are the ones that predict incremental revenue, and they are the ones that a well-configured AI loyalty agents platform should be reporting in real time.

Visit frequency uplift per cohort is the foundational metric. Segment your loyalty base by enrolment vintage (Month 0–3, 3–6, 6–12, 12+) and track whether visit frequency is increasing over time for each cohort. If it is not — if Month 6 members visit as infrequently as Month 1 members — the programme is not driving habitual behaviour, it is just capturing existing behaviour. Fundle's cohort analytics dashboard surfaces this automatically, flagging cohorts where frequency is flattening before they enter churn risk territory.

Incremental basket size is the second metric. Measure the average transaction value for loyalty members versus a matched control group of non-members with similar demographic and geographic characteristics. The match is important — naive comparisons of members versus non-members will overstate the loyalty effect because high-value shoppers self-select into programmes. In well-run AI-powered programmes, genuine incremental basket uplift of 18–28% is achievable over a 12-month programme horizon.

Redemption rate and redemption velocity matter because they measure perceived value. A redemption rate below 10% indicates that customers do not find the reward structure meaningful or accessible — a classic problem with point-expiry models used by EasyRewardz and older Capillary implementations. AI agents improve this by making redemption contextual and time-bound: the offer appears when the customer is present and ready to transact, not 30 days after the earning event.

Tenant revenue attribution is the metric that will get your CFO's attention. For each tenant brand, track the revenue generated by loyalty-driven visits versus organic visits. This becomes the commercial lever for tenant co-funding of the loyalty programme — if Pantaloons or Lifestyle can see that loyalty-driven customers generate ₹2,400 more per year than organic customers, they have a clear ROI for co-investing in the platform. Fundle Brand Loyalty module generates per-tenant attribution reports that make these conversations straightforward.

How Fundle solves this

Fundle was purpose-built for the structural complexity of Indian mall and enterprise retail — not adapted from a Western SaaS product or retrofitted from a points-ledger platform. The Fundle AI Platform operates as a full-stack loyalty and customer engagement infrastructure, unifying data ingestion, AI-driven decisioning, campaign execution, and retail media monetisation in a single architecture. Today, Fundle powers marketing across 123+ malls with AI-powered loyalty engagement solutions — a footprint that spans Tier-1 metro properties and Tier-2 growth markets, giving the platform unmatched exposure to the behavioural diversity of the Indian shopper.

At the core is Fundle AI Agents — goal-conditioned autonomous agents that continuously optimise for operator-defined business objectives without requiring human campaign intervention at the execution layer. Unlike MoEngage or WebEngage, which are excellent channel orchestration platforms but fundamentally require human-defined journey logic, Fundle Agentic AI generates, tests, and refines its own engagement hypotheses. The practical result: a mall marketing team of three people can run the complexity of what previously required a CRM team of twelve.

Fundle Mall Loyalty provides the coalition programme infrastructure — multi-tenant reward orchestration, unified member profiles, cross-brand redemption, and POS-level transaction capture across every tenant in the property. Fundle Brand Loyalty gives individual tenant brands their own analytics interface, offer management tools, and attribution reporting, so participation in the mall programme does not feel like surrendering brand autonomy. The two layers are architecturally separated but analytically integrated — the mall sees the aggregate picture, the brand sees its own slice, and neither needs to expose raw customer data to the other.

Fundle AI Workflow is the automation backbone that connects data signals to executed actions — managing the sequence, timing, channel selection, and offer personalisation logic for every customer interaction. For mall operators exploring retail media, Fundle Reach packages first-party audiences into targetable segments with in-store conversion measurement, turning the loyalty programme into an advertising revenue asset. Vineet Narang's founding vision was precisely this: that loyalty is not a discount mechanism but a data and media business that happens to improve customer experience as a side effect. For CRM heads and mall marketing directors evaluating next-generation platforms, the operational question is no longer whether to adopt an AI loyalty agents platform — it is how quickly you can get one deployed before your competition does.