Fundle.ai: The Best Customer Engagement Platform India Retailers Need Now

What Is Omnichannel Engagement and Why It Is Not Multichannel

Multichannel means being present on multiple surfaces: a store, an app, a website, WhatsApp, email. Most Indian retailers cleared that bar three years ago. Omnichannel means that every surface shares the same customer record in real time so that the experience is continuous, not repetitive. The distinction sounds semantic until you price out the cost of getting it wrong.

Consider Apollo Pharmacy, which operates over 6,000 outlets and a fast-growing e-pharmacy vertical. A customer who buys a diabetic supplement at a physical store and later searches for the same SKU on the app should not receive a new-customer welcome offer. They should receive a refill reminder calibrated to their average purchase interval, possibly bundled with a lab-test offer their purchase history suggests they need. That is omnichannel engagement: using the complete customer record to make the next interaction feel like a continuation, not a cold start.

In the Indian context, omnichannel execution has three mandatory layers that are rarely discussed together. First, identity resolution across POS, app, and web — because Indian shoppers are far more likely to transact under a mobile number than an email address, and the same person may register differently at each touchpoint. Second, consent management aligned with India's Digital Personal Data Protection Act (DPDP, 2023), which mandates explicit purpose-linked consent before any personal data is processed for marketing. Third, channel orchestration that respects how Indian consumers actually communicate: WhatsApp is primary, SMS is fallback, email is low-open, and app push is declining as notification fatigue increases.

Getting all three right simultaneously is what separates a genuine customer engagement platform India operators can scale from a collection of point solutions duct-taped together. Platforms like Capillary and MoEngage handle parts of this stack well, but their architectures were not built with Indian mall tenant ecosystems or the DPDP consent framework as first-class requirements. The gap is real, and it is measurable in points redemption rates, repeat visit frequency, and average transaction value lift.

Challenges in Indian Retail Omnichannel Execution

The five execution gaps that consistently undermine omnichannel programs in Indian retail are worth naming precisely, because each one requires a different fix and most vendors only address one or two.

Gap one: POS fragmentation. India's retail POS landscape is spectacularly diverse. A single mall may have tenants running Petpooja, POSist, GoFrugal, Wondersoft, and three proprietary enterprise ERPs simultaneously. Each system captures transaction data differently, exports in incompatible formats, and updates at different intervals. Building a unified customer record across these systems is an integration problem that most SaaS loyalty platforms underestimate by a factor of three in both time and cost. The result is that loyalty point balances are often 24–48 hours stale, which destroys in-store staff confidence in the program and teaches customers that points are fictional.

Gap two: Identity resolution at Indian scale. Indian shoppers share phone numbers (between family members), change numbers frequently (especially post-MNP), and rarely use the same email address across platforms. Standard match-and-merge algorithms built for Western markets, where email is the primary key, routinely produce duplicate records at rates of 20–30 percent in Indian datasets. Tanishq, for example, has reported customer databases where family members appear as a single high-value customer because all purchases were billed to one phone number. Engagement programs built on dirty identity data personalise to a ghost.

Gap three: Consent management under DPDP. The Digital Personal Data Protection Act, 2023 is not optional. Retailers who collected customer data via paper loyalty forms or checkbox opt-ins before the Act's notification are sitting on data whose legal basis for marketing use is uncertain at best. Any engagement platform deployed today must have a consent management layer that captures, stores, and honours granular purpose-linked consents and can produce an audit log on demand. This is not a feature most legacy CRM platforms were built to provide.

Gap four: Channel mis-sequencing. Most Indian retail CRM setups blast the same message on all channels simultaneously — SMS, email, push, and WhatsApp at once. This wastes budget and trains customers to ignore the brand. The correct approach is a priority waterfall: WhatsApp first (85 percent open rate in Indian retail contexts), push notification second for app-installed users, SMS as authenticated fallback, email for long-form content only. Platforms like Xeno and Almonds.ai are beginning to model this, but rule-based waterfalls are still common where AI-driven next-best-channel selection is needed.

Gap five: Mall-brand data silos. This is uniquely Indian and almost entirely unsolved by global platforms. A shopper who spends ₹4,500 at FabIndia and ₹2,800 at Cafe Coffee Day in the same Phoenix Marketcity visit is a highly valuable omni-tenant customer from the mall's perspective. But neither brand's loyalty system sees the full visit picture, and the mall has no mechanism to credit or reward that cross-tenant behaviour. The result is that the mall's loyalty program becomes a discount vehicle rather than a genuine engagement platform, and tenants continue to run parallel programs that cannibalise each other.

Fundle's Technology Stack for Unified Consumer Profiles

A unified consumer profile is not a database table. It is a living record that reconciles behavioural signals from a minimum of six data streams — POS transactions, app events, web sessions, Wi-Fi dwell data (for malls), social commerce interactions, and consent-verified survey responses — and keeps that record fresh enough to power real-time decisioning at the moment a customer enters a store or opens a message.

The Fundle AI Platform was architected around this requirement from day one. Its identity resolution engine uses a probabilistic graph model that weights mobile number, device fingerprint, and behavioural sequence similarity rather than relying on email as a primary key. In Indian retail deployments, this approach reduces duplicate customer records by 60–70 percent compared to standard CRM import processes, which translates directly into more accurate personalisation and fewer wasted communication events.

Fundle Loyalty sits on top of this unified profile layer and adds a rules-and-rewards engine that is explicitly designed for the mall-tenant relationship. A property using Fundle Mall Loyalty can configure a cross-tenant earn model — for example, ₹100 spent at any participating tenant earns 10 Fundle points redeemable anywhere in the mall — while simultaneously allowing each tenant to run brand-specific bonus campaigns through Fundle Brand Loyalty without data leakage between competing tenants. This is architecturally novel in the Indian market. Capillary's multi-tenant module and EasyRewardz's mall product both require data to be siloed at the tenant level, which means the mall never sees the cross-tenant spend picture that makes omnichannel engagement valuable.

Fundle AI Agents — the platform's conversational intelligence layer — handle the customer-facing side of this architecture. An AI agent can respond to a WhatsApp message asking 'how many points do I have?' in under two seconds, trigger a birthday offer when the profile flags an upcoming anniversary, or escalate a complaint to a human store manager with full context attached. Fundle Agentic AI goes further: it autonomously runs A/B tests on offer constructs, reallocates campaign budget toward higher-converting segments mid-flight, and generates weekly natural-language performance summaries for marketing heads who do not have time to read dashboards. Fundle AI Workflow is the orchestration layer that connects all of this to existing POS systems, ERPs, and communication APIs without requiring the retailer to rip and replace their current technology stack — a critical feature given the POS fragmentation reality described earlier.

KPIs That Separate Genuine Engagement from Vanity Metrics in Indian Retail

The Indian retail industry has a metric hygiene problem. Loyalty programs are routinely evaluated on enrolled member count and points issued — both of which are easy to inflate and tell operators almost nothing about economic value created. A Pantaloons store that enrolled 80,000 members in a single quarter through a cash-discount promotion has not built an engagement asset. It has built a coupon database. The distinction matters because the former generates incremental margin and the latter erodes it.

The six metrics that actually predict program health in Indian retail contexts are as follows. First, redemption rate: the percentage of points issued that are actually redeemed within 180 days. Industry benchmarks suggest healthy programs in Indian fashion and lifestyle retail achieve 22–35 percent redemption rates. Rates below 15 percent indicate that customers do not value the reward construct or do not understand how to redeem. Rates above 40 percent may indicate over-generous earn ratios that will compress margins.

Second, repeat visit rate at 30 and 90 days: what percentage of customers who transacted once return within 30 days and again within 90 days? In Indian mass-market apparel (Reliance Trends, Lifestyle), the 30-day repeat rate for non-loyalty members averages 12–15 percent. Well-run loyalty programs should push this to 22–28 percent. If your program is not moving this number, the engagement content is not compelling enough to change behaviour.

Third, average transaction value lift in loyalty vs. non-loyalty cohorts: loyalty members should transact at 15–25 percent higher basket values than non-members in the same store, after controlling for self-selection bias. If the lift is below 10 percent, the program is attracting your existing best customers rather than upgrading average ones.

Fourth, message opt-out rate: any campaign generating opt-outs above 2.5 percent per send is over-communicating or mis-targeting. Indian shoppers are highly sensitive to irrelevant messages, particularly on WhatsApp where unsubscribes carry social friction. Fifth, cost per engaged customer: total program cost divided by customers who both transacted and redeemed in the measurement period. This is your true unit economics figure. Sixth, CSAT and NPS delta between loyalty and non-loyalty cohorts: loyalty members who feel genuinely recognised should score 8–12 NPS points higher than non-members. If they do not, the personalisation is not working.

How Fundle solves this

Fundle was built on a single conviction: that Indian retail operators deserve an AI-native customer engagement platform that treats the mall-brand-customer triangle as the fundamental unit of loyalty architecture, not an edge case. Every layer of the Fundle AI Platform reflects this conviction in its data model, its consent management, and its channel orchestration logic.

Fundle Loyalty provides the foundational earn-and-burn engine, but unlike legacy points platforms, it is designed for multi-operator environments from the start. A mall running Fundle Mall Loyalty can onboard 50 tenants in a single property and give each tenant access to their own customer cohort through Fundle Brand Loyalty without exposing competitor data. The mall operator sees the full cross-tenant spend picture; each brand sees only its own customers plus anonymised benchmarks. This data-sharing architecture is what allows Fundle to synchronize loyalty and engagement data across offline and online channels for 270+ Indian brands — a fact that reflects both the platform's technical depth and its commercial trust model.

Fundle AI Agents power the customer-facing conversational layer across WhatsApp, app, and web. They handle tier-status queries, points redemption, complaint routing, and personalised offer delivery without human intervention, at a cost per interaction that is 80–90 percent lower than an equivalent call-centre resolution. Fundle Agentic AI takes autonomous action on behalf of the marketing team: it identifies a cohort of customers who visited the mall twice in the last 30 days but have not transacted at a high-margin tenant, builds a targeted cross-sell offer, tests three creative variants, selects the winner at statistical significance, and scales — all without a human writing a campaign brief. Fundle AI Workflow is the integration backbone that makes this possible without a six-month IT project: pre-built connectors for POSist, Petpooja, GoFrugal, Wondersoft, and major Indian ERP platforms mean that most deployments go live in 8–12 weeks.

Vineet Narang's founding vision for Fundle was grounded in a frustration he observed repeatedly across Indian retail: that the technology stack serving loyalty and engagement had not evolved to match the complexity of how Indian consumers actually shop. The platform he built addresses that gap directly — with AI that understands Indian channel preferences, a consent engine designed for DPDP from the ground up, and an identity resolution model that handles the realities of Indian customer data rather than assuming Western norms. For any retail marketing head, mall CMO, or loyalty manager who has spent years watching point-based programs fail to move repeat visit rates, Fundle's architecture represents a fundamentally different approach — one that is measurable, DPDP-compliant, and built for the Indian market at Indian scale.