India’s digital pull revolution: How kirana stores are reshaping global retail

Growing up in India, I walked to the kirana store — a small neighborhood grocery — almost every day for milk, cooking oil, lentils and, in the summer, daily ice creams. Now, living in the United States, I plan a biweekly grocery run because the nearest store is a drive away, not a walk. The difference is not just convenience; it is a fundamentally different model of how retail works.

Walk into any lane in an Indian town and you will find it: a store barely the size of a living room, stacked floor to ceiling, run by a single owner who knows almost every customer by name. India’s retail landscape is largely unorganized and in 2021 was made up of roughly 13 million kirana stores, which contribute nearly 11% of the country’s gross domestic product (GDP) and 8% of its total workforce.

For decades, the Indian fast-moving consumer goods supply chain operated through a rigid, multilayered chain: manufacturer to super stockist to distributor to kirana owner and finally to the consumer. With 12 million grocery retail outlets and 1 million wholesalers and distributors, the Indian grocery retail ecosystem is one of the most complex in the world.

At the heart of this complexity was a fundamental gap in how demand was understood: The only demand signal available was the salesman’s weekly visit. Orders were placed based on gut feeling, personal relationships and whatever the distributor wanted to push, not what the store needed. The result was a paradox that plagued every link in the chain simultaneously: too much of the wrong product sitting unsold, too little of the right one on the shelf.

The fragility of this opaque, cash-heavy system was fully exposed when COVID-19 hit. Demand for many items became violently volatile almost overnight, while lockdowns slowed or interrupted the physical flow of materials across the country, revealing that a distribution network built entirely on human relationships and gut instinct had no mechanism to absorb shock. This was not a broken system so much as a system that was never designed with data in mind.

The push model vs. the pull revolution

The American distribution model was built on different logic: Produce in bulk; push inventory through consolidated warehouse networks; and let forecast-driven algorithms determine what reaches the shelf. This model delivered extraordinary efficiency at scale, but its central vulnerability was complete dependence on accurate demand forecasting. When manufacturing facilities shut down during the pandemic, shelves emptied not because goods did not exist, but because a warehouse-centric push system had no mechanism to reroute in real time.

The structural contrast with India is stark: The United States has approximately 1.07 million total retail establishments, of which true large-format stores — Walmart, Target, Costco, and similar chains — number in the tens of thousands. For context, Walmart operates around 5,300 U.S. stores, Kroger around 2,800, and Target around 2,000. Even adding every major chain together, large-format retail accounts for well under 100,000 locations nationally. India’s distribution, on the other hand, runs through 12 million stores, the vast majority under 500 square feet, where product availability at the neighborhood level is the primary driver of sales. The systems represent two philosophies — two architectures — and the pandemic exposed which one breaks first.

India’s digital pull revolution

The shift began quietly, a kirana owner in a small town opening an app at 10 p.m., placing an order based on what sold that day, and receiving delivery by morning. No salesman was directly involved, no guesswork, and no dead stock. This is the digital pull revolution now reshaping how consumer goods move through India.

Hindustan Unilever’s Shikhar app, launched as a mobile platform allowing kirana stores to place orders directly 24 hours a day, changed the demand signal: Instead of waiting for a sales representative’s weekly visit, store owners could instantly order when stock ran low, with orders often fulfilled the next day. Shikhar now accounts for almost one-third of Hindustan Unilever’s sales from neighborhood retailers, a metric that would have been unimaginable in the manual, relationship-driven system it replaced.

On the business-to-business side, platforms like Udaan replicated this logic at national scale. Udaan’s reach now extends beyond 1,200 cities in India for daily delivery and over 12,500 postal codes through its supply chain system, connecting 3 million retailers to thousands of sellers across consumer goods, grocery, pharmaceuticals, and lifestyle categories. Through its Project Vistaar, Udaan plans to expand sixfold to reach around 10,000 small towns and villages — markets that the traditional distributor model considered unreachable.

The numbers that matter

The case for digital transformation is not theoretical. According to a study by Accenture, every kirana store that went through digital transformation experienced revenue growth in the range of 20 to 300 percent, while profits grew by 30 to 400 percent. The primary driver was not selling more; it was wasting less. Dead stock was not eliminated by selling harder; it was eliminated by ordering smarter.

The scale of this shift is captured in three numbers:

• 70% of kirana stores in major cities expressed willingness to adopt technology to manage their businesses, according to RedSeer Consulting, compared to just 3% that were tech-enabled as of 2018.
• Just over 90% of fast-moving consumer goods sales in India still happen via kirana stores — a share that actually grew during the pandemic, according to CB Insights citing Nielsen data.
• Transforming just 10% of India’s 13 million kirana stores could boost retail consumption by more than 5% and generate approximately 3.2 million new jobs, according to Accenture and the Trust for Retailers and Retail Associates of India.

The resilience argument completes the picture. When COVID-19 disrupted centralized warehouse systems globally, India’s kirana network — hyperlocal, community-rooted, and operating on daily demand signals — held. One could argue that the resilience and the pull model led to the scenario where larger stores struggled to deliver, while consumers depended on kirana stores for daily essentials; due to their links with local distributors, these stores were able to meet demand where centralized formats could not.

The bigger lesson: Reverse innovation

The kirana story is not just an India story; it is a case study in what scholars Vijay Govindarajan and Chris Trimble termed “reverse innovation.” First articulated in a 2009 Harvard Business Review article, reverse innovation describes any innovation likely to be adopted first in the developing world before flowing upward into mature economies.

The kirana’s digital pull system emerged from the constraints of a small store with razor-thin working capital and no margin for error — constraints that, once met with the right technology, forced an efficiency the warehouse model never had to discover. Yet that constraint produced a model that eliminated dead stock, reached markets the formal economy had written off, and survived a global pandemic better than centralized warehouse systems did.

In the United States, the structural equivalent already exists. According to the Bodega Association of the United States, as reported by National Public Radio, as of 2020, New York City alone had an estimated 13,000 bodegas — small independent stores serving dense urban neighborhoods — that large-format retail has never fully penetrated. The tools exist. The demand exists. The digital pull model exists. What is missing is the willingness to look south for the playbook.

Can the digital pull model work beyond India?

The short answer is: It already does, in pockets. McKinsey’s analysis of digital business-to-business disruption in fragmented retail identifies parallel platforms operating across emerging markets. Meicai and Alibaba’s Lingshoutong in China; Bukalapak in Indonesia; and Chiper and Quqo in Latin America all operate on a structurally similar logic: replace the human salesperson’s weekly visit with an app-based ordering system that generates real-time demand signals. What they share is the underlying condition that makes the model viable: retail fragmentation. The McKinsey report estimates the addressable market for such platforms at over $2.8 trillion worldwide, precisely because fragmented trade is the dominant retail structure across most of Asia, Africa and Latin America. In Africa, the picture is still early-stage; the continent’s e-commerce market was projected to reach $40.5 billion in revenue in 2025, growing at a compound annual rate of around 8.5% through 2029, according to Statista, but the structural conditions mirror India’s: dense urban trade concentrated in small stores; high mobile penetration; and an entrenched cash economy now being unlocked by digital wallets.

The question is not whether the model is transferable. It is whether the enabling conditions — a critical mass of smartphone-equipped retailers, a reliable logistics layer, and a payment system capable of handling small-value transactions at scale — exist or can be built.

What stands in the way?

The barriers are more human than technical. A 2025 survey by BeatRoute of over 100 senior executives in India’s fast-moving consumer goods sector found that 42% of leaders reported friction with existing channel partners when introducing digital platforms, while 49% cited intensifying competition between traditional and digital channels.

The core tension is that distributors who have spent decades building retailer relationships do not give up that position quietly. In 2022, distributors in Maharashtra, a state in India, went on strike, refusing to supply products from Hindustan Unilever after the company’s direct-to-retailer app began bypassing them. This is actually a structural feature of any model that redistributes power from intermediary to brand or platform. Beyond stakeholder friction, infrastructure gaps remain a real constraint: Last-mile logistics in rural areas, inconsistent cold-chain coverage, and limited digital literacy among older kirana owners all slow adoption. Latin America faces delivery reliability and payment fraud challenges; Africa is still building the logistics backbone; and Southeast Asia contends with regulatory fragmentation and inconsistent internet quality across borders, according to market analysis by IMARC Group.

What technologies are actually doing the work?

Four layers of technology underpin the transformation. The first is the mobile ordering interface itself — low-friction apps that work on basic Android devices and require minimal data bandwidth.

The second is the payments layer: India’s Unified Payments Interface (UPI) processed 228.3 billion transactions worth approximately $3.6 trillion across 2025, according to data from the National Payments Corp. of India, making it the world’s largest real-time payment system by volume, surpassing Visa. In December 2025 alone, UPI recorded 21.63 billion transactions, and its near-ubiquity makes instant, small-value digital transactions economically viable at the kirana level. Without a payment rail that works for a $2 transaction, the ordering app is useless.

The third layer is artificial intelligence-driven demand forecasting — what Palantir, SAP and proprietary platforms like Udaan use to translate transaction data into predictive inventory recommendations, reducing dead stock at the store level and improving supply planning accuracy across the chain.

The fourth and perhaps most structurally significant is India’s Open Network for Digital Commerce, known as ONDC, a government-backed open protocol that functions like a UPI for commerce, allowing any buyer or seller app to connect to a shared network without being locked into a proprietary platform like Amazon or Flipkart. McKinsey estimates that ONDC has the potential to increase digital consumption in India by $340 billion by 2030. By November 2024, the platform was processing 14.45 million monthly transactions, and Amazon is already working with ONDC to digitize kirana stores. The combination of these four layers makes the Indian model replicable rather than a one-off.

What comes next?

The next wave is already visible in India. Quick commerce — 10- to 30-minute delivery through hyperlocal dark stores — is pushing the pull model to its logical extreme: demand fulfilled not in hours, but in minutes. Blinkit alone had crossed 1,544 dark stores as of June 2025, as per the first-quarter Zomato (now Eternal Ltd.) shareholder filing for fiscal year 2026 (July 2025), and India’s total dark store network stood at approximately 2,525 across more than 100 cities as of October 2025, according to Storyboard18. Across India’s gig economy, NITI Aayog estimates that approximately 6 million workers were engaged in last-mile delivery roles as of fiscal year 2025, with the total gig workforce expected to reach 23.5 million by 2030. But the more durable trend is the integration of artificial intelligence into every layer of the retail stack.

India Brand Equity Foundation’s January 2026 analysis of quick commerce found that companies are already deploying AI-based demand forecasting and dynamic route optimization to reduce waste and improve delivery speeds, while warehouse automation and micro-fulfillment centers are being piloted across major cities. ONDC’s road map includes AI-driven cataloging in regional languages and multilingual voice-enabled interfaces, meaning a kirana owner in rural Bihar could eventually manage inventory through a voice command in Bhojpuri. Looking further out, the systemic risk in this ecosystem is also becoming clearer: As platforms deepen their hold on the retailer relationship, the kirana risks becoming a fulfillment node rather than an independent business. That tension between the upgrade the digital model offers and the dependency it creates is the unresolved question that will define India’s retail future and eventually the retail future of every market that follows this path.

The kirana is not being disrupted; it is being upgraded

Bain & Co.’s 2025 Consumer Products Report indicates that emerging markets experienced an 11% rise in retail sales value in 2024. This is more than double the rate in developed markets. Emerging markets were responsible for all the volume growth recorded globally that year. These developing markets require assistance. A model that works for stores in developed markets cannot serve them. Emerging markets require retail establishments in proximity to residential areas. They have to identify what consumers are interested in purchasing. The digital pull paradigm serves as a method to do this. It operates by analyzing real purchasing behavior rather than relying just on computer-generated predictions. For emerging markets, this model is ideal. This concept was not invented in India. It was the first to succeed. India had a large population but little wealth. It was necessary for this model to function. Leaders in consumer products no longer have to decide whether to use the digital pull paradigm. It is the speed at which the platform, rather than the brand, takes control of the retailer relationship. The issue was never with the kirana. It was the lack of data. The 100-square-foot store that previously relied on memory and trust now works on both, and it turns out that this combination is harder to disrupt than any warehouse.

About the author

Anuprita Kaple earned her Master of Engineering in Engineering Management at Cornell University in 2026 and participated in the Samuel Curtis Johnson Graduate School of Management’s strategic product and marketing immersion, which has consumer goods as a primary focus. She has a background in global supply chain management and engineering operations. Prior to Cornell, she worked in digital transformation, overseeing cross-functional strategy operations across Europe, North America, emerging markets and China. She holds a bachelor of technology in mechanical engineering and was a fellow of the Cañizares Center for Emerging Markets in 2025-26.

All views expressed in articles published on the Cañizares Center for Emerging Markets webpage are those of the author(s) and should not be taken as reflecting the views of the Cañizares Center for Emerging Markets.