“We didn't build Fundle to sell software. We built it to make first-party data productive — every campaign, every store, every shopper, every day.”
- •Understand why India's loyalty market is entering a structural inflection point driven by UPI, ONDC, and first-party data mandates
- •Identify the five AI trends reshaping how Phoenix Marketcity-scale operators and multi-brand chains run campaigns
- •Assess the gap between point-based legacy programs and predictive, agentic campaign orchestration
- •Follow a five-step strategic roadmap to migrate to AI-driven loyalty campaign management without disrupting existing POS stacks
- •Benchmark your program against Fundle's tracked ₹2,329Cr+ revenue ecosystem and adapt continuously
India's organised retail sector crossed ₹11.5 lakh crore in FY24 and is on track to become the world's third-largest retail market by 2030. Yet if you walk the floors of any Grade-A mall — Select CITYWALK in Saket, Phoenix Marketcity in Pune, or Nexus Seawoods in Navi Mumbai — you will find marketing heads running the same playbook they ran in 2017: a points-per-rupee program bolted onto a POS, a monthly SMS blast, and a birthday coupon that costs more to redeem than it earns in incremental revenue. The tools have upgraded marginally. The thinking has not.
The argument for AI-driven loyalty campaign management India is not about technology novelty. It is about a structural gap that is widening fast. India now has 750 million smartphone users, 330 million UPI MAUs, and a rapidly maturing D2C cohort that has trained consumers to expect hyper-personalised engagement. Brands like Tanishq, Manyavar, and Lenskart have set a personalisation benchmark at the brand level. Mall operators and multi-brand chains are scrambling to match it at the portfolio level — across 80 to 200 tenants, each with its own SKU universe, margin profile, and customer segment. Legacy systems were never architected for that complexity.
The data problem compounds the campaign problem. Traditional loyalty platforms — whether on-premise installs running on GoFrugal or Wondersoft, or first-generation SaaS stacks — capture transactions but not intent. They tell you what a customer bought; they cannot tell you why she has not returned in 47 days, which tenant she is most likely to visit next, or at what discount depth she will reactivate without cannibalising margin. Answering those questions at scale, across thousands of members, in near-real-time, is precisely what AI-driven campaign orchestration is designed to do.
This is the context in which Fundle was built. The platform is purpose-designed for Indian mall operators and enterprise retail chains that need to move from reactive, calendar-driven campaigns to predictive, continuously-learning engagement engines. The stakes are concrete: programs that move from batch campaigns to AI-orchestrated journeys typically see 18–25% improvement in repeat visit frequency and 12–20% lift in average transaction value within 12 months of deployment. This article maps the landscape, the trends, the regulatory shifts, and the playbook — so that India's retail marketing leaders can act, not just plan.
India Retail Loyalty: The Numbers That Matter in 2025
Evolving Retail Loyalty Landscape in India
The Indian loyalty market has gone through three distinct phases in the past decade. Phase one was pure points accumulation — programs modelled on airline miles, ported to retail, and almost entirely transactional. Phase two was digital-first enrollment: mobile apps, QR codes at checkout, push notifications. Most major mall operators and chains — Lifestyle, Pantaloons, Reliance Trends — built or bought these capabilities between 2016 and 2021. Phase three, which is where the leading operators are today, is about moving from engagement to intelligence: using behavioural data to predict, personalise, and automate.
What is driving the transition is not ambition alone — it is competitive pressure from three directions. First, D2C and quick-commerce brands have raised the bar on personalisation speed. A Nykaa or Blinkit can run a segment-of-one campaign within hours of a trigger event; a mall retailer running on a legacy loyalty stack takes three to five days to conceptualise, approve, and deploy a similar campaign. Second, UPI and account aggregator frameworks are making cross-merchant transaction data more accessible, meaning the brands that build first-party data infrastructure now will have a durable moat when aggregated data becomes table stakes. Third, the maturation of the Indian consumer — particularly the ₹5–25 lakh annual household income segment — means that a flat 2% cashback is no longer differentiating. Emotional loyalty, experience-based rewards, and surprise-and-delight mechanics are now the vectors of differentiation.
For mall operators specifically, the challenge is structural. A Phoenix Marketcity property might have 180 tenants spanning F&B, apparel, electronics, and entertainment. Each tenant has its own loyalty or CRM system — Tanishq runs Encircle, FabIndia has its own membership, Cafe Coffee Day runs Brew Bucks. The mall operator sits on top of this fragmented stack trying to run a unified member journey with no single view of the customer. AI-driven loyalty campaign management India solves this not by replacing tenant-level programs but by creating a federated intelligence layer that can orchestrate cross-category campaigns without requiring tenants to surrender their proprietary data.
The implication for marketing heads is clear: the loyalty program of 2025 is not a points ledger. It is a continuously learning engagement engine that reads signals, predicts next-best actions, and deploys campaigns through the right channel at the right moment. The operators who understand this are already building it. The ones who do not will find their footfall numbers and basket sizes diverging from market averages within 18 months.
The AI-Driven Loyalty Campaign Funnel: From Signal to Revenue
Emerging AI Trends Impacting Loyalty Campaigns
Three AI trends are converging to fundamentally reshape how Indian retail marketing heads should think about campaign management in the next 24 months. Understanding each trend at an operator level — not just at a vendor pitch level — is what separates strategic capital allocation from technology spend with no measurable return.
The first trend is agentic AI for campaign orchestration. Traditional marketing automation — the kind offered by MoEngage, WebEngage, or first-generation Capillary configurations — operates on rule trees. If a customer has not visited in 30 days, send an SMS. If the SMS is opened but not redeemed, send a push notification with a bigger offer. Agentic AI is categorically different: an AI agent monitors multiple signals simultaneously, forms a hypothesis about why a member is disengaging, selects from a library of campaign tactics, tests variants autonomously, and reallocates budget to the highest-performing variant — all without a human setting up each rule. This is what Fundle Agentic AI is architected to do: not automate human decisions but replace the decision loop entirely for high-frequency, low-stakes campaign choices, freeing marketing teams to focus on strategy and creative.
The second trend is real-time RFM decay detection. The Recency-Frequency-Monetary framework has been a retail staple for 30 years, but most Indian retailers run RFM analysis monthly or quarterly in batch mode. AI-native platforms can now detect RFM decay at the individual member level in near-real time — identifying the exact moment a previously high-frequency shopper misses their expected visit window and triggering a reactivation campaign before the disengagement becomes habitual. Apollo Pharmacy, for instance, operates in a category where a customer who typically refills a chronic medication every 28 days but has not appeared on day 32 is at high risk of switching to a competitor. Catching that signal on day 33, not day 60, is the difference between retention and churn.
The third trend is generative AI for offer personalisation. Most loyalty platforms today send the same offer architecture to an entire segment: 10% off at any F&B outlet this weekend. Generative AI models trained on transaction history, category affinity, and price sensitivity can now draft individualised offer narratives — specific outlet recommendations, personalised discount depths, and reward mechanics calibrated to each member's redemption history. This is not science fiction; Xeno and Almonds.ai have early versions of this in market. The competitive question for mall operators and chains is not whether this capability exists but whether their data infrastructure is clean enough to power it. Garbage in, garbage out — and most Indian loyalty databases have significant data quality problems that need to be resolved before AI personalisation can function at scale.
The fourth trend, less discussed but equally important, is AI-powered campaign ROI attribution. Indian retail has historically struggled with closed-loop attribution — connecting a campaign send to an in-store transaction without relying on coupon redemption as a proxy. New approaches combining device fingerprinting, WiFi analytics, UPI transaction matching, and loyalty card swipes are enabling genuine multi-touch attribution models. This matters enormously for marketing budget justification: a CMO who can show that a WhatsApp campaign generated ₹47 in incremental revenue per rupee spent has a fundamentally different conversation with the CFO than one presenting open-rate statistics.
Legacy Loyalty Platforms vs. AI-Driven Campaign Management: Operator Reality Check
Preparing for New Data Privacy Regulations in Indian Retail
The Digital Personal Data Protection Act 2023 (DPDPA) is not a future concern — it is an active operational requirement that Indian retail marketers need to have addressed in their loyalty architecture today. The Act mandates explicit, purpose-limited consent for data collection, grants consumers the right to erasure and data portability, and imposes significant penalty exposure for non-compliant processing. For a loyalty program that collects transaction history, location data, communication preferences, and potentially health or financial data (Apollo Pharmacy, Tanishq, for instance), the compliance surface is substantial.
The practical implication for loyalty program operators is threefold. First, your enrollment flow must collect granular consent at the purpose level — consent to receive promotional SMS is legally distinct from consent to use transaction data for AI-based profiling. Programs that have a single opt-in checkbox at enrollment are already non-compliant and need re-architecture. Second, your data infrastructure must support the right-to-erasure workflow: if a member requests deletion, you need to be able to purge their data from your loyalty database, your campaign management system, your data warehouse, and any third-party integrations — within the timeframes the Act specifies. Third, cross-brand data sharing in a mall context — where the operator might share a member's footfall pattern with tenant brands to enable joint campaigns — requires explicit bilateral consent that most current mall loyalty programs do not obtain.
This is where an AI-native loyalty platform architecture has a structural advantage over legacy stacks. A platform like Fundle Mall Loyalty that was built after the DPDPA framework was published can embed consent management, purpose-limitation controls, and erasure workflows natively into the data model. A legacy platform that was architected before 2022 and has consent bolted on as a module is operationally fragile — consent records live in a separate system from the campaign engine, creating reconciliation gaps that are both compliance risks and operational headaches.
Beyond DPDPA, the shift away from third-party cookies and MAID-based mobile targeting is accelerating the first-party data imperative. Google's Privacy Sandbox rollout and Apple's ATT framework have already degraded the effectiveness of paid re-engagement for retail apps by 30–40% in measurable CPM terms. The loyalty program is, increasingly, the primary first-party data collection mechanism for Indian retailers. Marketing heads who treat it as a promotions vehicle rather than a data asset are making a category error that will compound over the next three years.
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.
Strategic Roadmap: Five Steps to AI-Driven Loyalty Campaign Management
Data Infrastructure Audit and Cleanse
Before any AI model can function, the underlying member database must be accurate. Audit for duplicate member IDs, missing phone-email links, stale contact data, and inconsistent transaction tagging across POS systems (GoFrugal, Petpooja, POSist). Indian retail databases typically have 25–40% data quality issues. Allocate 6–8 weeks for this phase. No AI campaign engine performs well on dirty data — this is the unglamorous prerequisite that most technology vendors underplay.
Consent Architecture Redesign for DPDPA Compliance
Redesign your enrollment and preference centre to capture granular, purpose-specific consent. Map every data element you collect to a specific business purpose. Build erasure and portability workflows before they become a regulatory demand. This step is best done in parallel with the data audit, not after, because consent gaps will affect what data you can legally use in AI models.
AI Segmentation and Propensity Model Deployment
Deploy RFM-based segmentation as the baseline, then layer propensity models: next-category-purchase likelihood, churn-risk score, discount sensitivity index, and cross-tenant visit propensity for mall operators. Initial models should be trained on 12–18 months of historical transaction data. Expect the first 60 days to be model calibration — do not judge AI campaign performance in the first month.
Omnichannel Campaign Orchestration Integration
Connect your AI campaign engine to every outbound channel: WhatsApp Business API, SMS, push notifications, email, and in-mall digital screens where available. Map the right message type to the right channel based on member behaviour — high-engagement members respond to WhatsApp; lapsed members often need an SMS with a hard offer. Set up control groups for every campaign to maintain clean incrementality measurement.
Closed-Loop Attribution and Continuous Learning Setup
Close the loop between campaign send and in-store transaction using POS integration, QR-based redemption tracking, and UPI reference matching where available. Feed attribution signals back into the AI model on a weekly basis. Define your primary KPIs — incremental revenue per campaign, reactivation rate, repeat visit frequency — and review them in a weekly campaign performance ritual, not monthly. The AI improves fastest when the feedback loop is tight.
Integrating AI with Omnichannel Retail Experiences
The word omnichannel has been so overused in Indian retail that it has nearly lost operational meaning. For the purposes of AI-driven loyalty campaign management India, omnichannel means one specific thing: the ability to recognise the same member across every touchpoint — in-store POS, mobile app, website, WhatsApp conversation, and in-mall WiFi session — and orchestrate a coherent, non-repetitive engagement experience across all of them. Most Indian retailers are not there yet, and the gap is not primarily technical. It is organisational: the app team, the CRM team, and the store operations team each own a slice of the customer data and protect it jealously.
The technical integration challenge is real but tractable. A typical multi-brand retailer running Lifestyle or Pantaloons stores might have POSist at the checkout, a custom mobile app, a WhatsApp Business API integration managed by a different agency, and email running on a legacy ESP. Each of these systems has a member identifier — phone number, email, loyalty card number, device ID — but they are rarely matched into a single unified profile. Building that profile unification layer is the foundational technical task, and it is where investments in a Customer Data Platform (CDP) or an AI-native loyalty platform with built-in identity resolution pay off quickly.
For mall operators, the integration challenge has an additional dimension: tenant data sharing. A Fundle Mall Loyalty deployment at a Phoenix Marketcity property needs to ingest transaction signals from 150+ tenants, each running different POS systems — some on POSist, some on GoFrugal, some on proprietary stacks. The integration approach must be API-first and POS-agnostic, with a fallback to QR-based transaction capture for tenants whose POS systems cannot expose transaction APIs. This is an engineering detail that has major implications for campaign quality: a mall loyalty program that captures only 60% of member transactions cannot build accurate propensity models.
The omnichannel payoff, when the integration is done correctly, is substantial. Members who are engaged across three or more channels — app, WhatsApp, and in-store — show 2.4× higher annual spend than single-channel members in comparable Indian retail deployments. The AI campaign engine's job is to find every member's highest-engagement channel combination and optimise communication accordingly, not to blast every member on every channel and call it omnichannel. The latter approach burns notification permissions and increases opt-out rates; the former compounds engagement over time. This distinction is where AI-powered customer engagement campaigns genuinely separate from traditional marketing automation.
- Member database has less than 15% duplicate or unmatchable records, with phone number as the primary unique identifier across all POS integrations
- Enrollment flow captures granular, purpose-specific consent compliant with DPDPA 2023 — not a single blanket opt-in
- Transaction data from all POS systems (GoFrugal, POSist, Petpooja, Wondersoft) flows into a unified member profile within 15 minutes of purchase
- RFM segmentation is refreshed at minimum weekly; churn-risk scoring runs on a daily batch or real-time trigger basis
- At least two outbound channels (WhatsApp + SMS, or app push + email) are live with campaign orchestration, and channel preference is tracked per member
- A control group methodology is in place for all campaigns to distinguish incremental lift from organic behaviour — without this, AI campaign ROI cannot be credibly measured
- A named campaign performance owner reviews AI-generated campaign results weekly and feeds qualitative context (promo calendar, competitor activity, seasonal shifts) back to the AI model configuration
“Indian retail has enough data to win loyalty wars — what it lacks is the intelligence layer that turns transaction records into predictive actions. That gap is exactly what we built Fundle to close.”
How Fundle solves this
Fundle was purpose-built for the specific complexity of Indian retail loyalty — not adapted from a Western enterprise CRM or a generic marketing automation stack. The Fundle AI Platform sits at the intersection of transaction intelligence, member engagement, and campaign orchestration, with an architecture that is explicitly designed for the multi-tenant, multi-POS, multi-channel reality of Indian malls and enterprise retail chains. Fundle tracks ₹2,329Cr+ revenue and continually adapts campaign AI to India's evolving retail ecosystem — a live production signal that informs how its models are calibrated, not a sandbox benchmark.
For mall operators, Fundle Mall Loyalty provides the federated intelligence layer described earlier: a unified member profile built from cross-tenant transaction signals, with consent-aware data sharing controls that allow the mall operator and participating tenants to run joint campaigns without violating member privacy or DPDPA obligations. The platform ingests POS data from GoFrugal, Petpooja, POSist, and Wondersoft through pre-built connectors, reducing integration timelines from months to weeks. RFM decay detection runs continuously, and Fundle Agentic AI generates campaign recommendations — including channel, offer depth, and timing — that marketing teams can approve with a single click or configure to run fully autonomously for pre-approved campaign types.
For enterprise retail chains operating Fundle Brand Loyalty — think a Manyavar, FabIndia, or a regional grocery chain with 200+ stores — the Fundle AI Workflow engine manages the full campaign lifecycle: from segment identification through offer generation, channel dispatch, redemption tracking, and closed-loop attribution reporting. The Fundle AI Agents layer handles high-frequency micro-decisions — which of 400 lapsed members to re-engage this Tuesday, at what offer depth, through which channel — freeing the four-person CRM team to focus on quarterly strategy and creative briefs rather than daily campaign configuration.
Vineet Narang's founding thesis was that Indian retail deserved a loyalty platform that was as intelligent as the market is complex — one that could handle the INR-denominated, UPI-native, DPDPA-compliant, multi-lingual reality of 1.4 billion consumers across 29 states without requiring a Fortune 500 technology budget to operate. The Fundle AI Campaign Management capability reflects that thesis in production: AI-powered customer engagement campaigns that are genuinely accessible to a 50-store regional chain and powerful enough for a 200-tenant Grade-A mall. The competitive set — Capillary, EasyRewardz, Customer Capital — offers capable platforms, but none were architected from inception for the agentic AI campaign management model that Fundle Agentic AI delivers. That architectural difference compounds over time as AI models improve with more data and more feedback loops — making the decision to move to an AI-native platform now, rather than in two years, a question of compounding advantage versus compounding catch-up.
Frequently asked
What is AI-driven loyalty campaign management and how is it different from traditional marketing automation?+
Traditional marketing automation executes rule trees: if X, then Y. AI-driven loyalty campaign management uses machine learning models to predict member behaviour, generate personalised offers, select optimal channels and timing, and continuously improve based on campaign outcomes — without requiring manual rule configuration for every scenario. The practical difference is speed and scale: an AI system can manage thousands of individualised campaign decisions simultaneously, while a rule-based system requires a human to anticipate and configure every decision path in advance.
How does DPDPA 2023 affect Indian retail loyalty programs specifically?+
DPDPA requires explicit, purpose-limited consent for every type of data processing. For loyalty programs, this means separate consent for transaction tracking, AI-based profiling, cross-brand data sharing, and promotional communications. Programs must also support right-to-erasure requests across all systems — including the campaign engine, data warehouse, and any third-party integrations. Mall operators sharing member data with tenants for joint campaigns need bilateral consent architecture that most current programs do not have.
Can AI campaign management work for smaller retail chains with limited data history?+
Yes, with calibrated expectations. Models trained on fewer than 50,000 transaction records will produce wider confidence intervals and require more conservative campaign strategies initially. The practical approach is to start with RFM-based segmentation — which functions with as few as 6 months of transaction history — and layer propensity models as the data volume grows. A 30-store regional chain with 18 months of clean POS data can run effective AI-assisted campaigns; the models simply improve faster as the member base grows.
How long does it typically take to see ROI from an AI-driven loyalty platform deployment in India?+
The honest answer is 3–4 months to first measurable signal, 6–9 months to stabilised ROI. The first two months are infrastructure: data cleanse, POS integration, consent architecture. Month three is baseline campaign calibration. From month four, you typically see measurable lift in reactivation rates and average transaction value for AI-triggered campaigns versus control groups. Full ROI — including the cost of migration from legacy systems — is typically recovered within 12–15 months for mid-size operators.
How does Fundle handle multi-tenant data sharing for mall loyalty programs?+
Fundle Mall Loyalty uses a federated data model: each tenant's transaction data is stored with tenant-level access controls. The Fundle AI Platform aggregates signals at the member level for campaign orchestration purposes only, with consent flags governing which data elements can be shared with which tenant parties. Joint campaign reports show incremental footfall and basket lift without exposing individual member-level transaction data to tenant brands — maintaining both privacy compliance and competitive confidentiality between tenants.
What POS systems does Fundle integrate with natively?+
Fundle has pre-built API connectors for the major Indian retail POS stacks including Petpooja (F&B), POSist (restaurant and QSR), GoFrugal (grocery and fashion), and Wondersoft (apparel and lifestyle). For tenants or stores running proprietary or less common POS systems, Fundle provides a QR-based transaction capture fallback that maintains data continuity without requiring custom POS integration. The API-first architecture means new POS connectors can be added without platform-level changes.
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 Indian retail members, malls and 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.