City context
Varanasi is not a city that explains itself quickly. It is, by credible historical reckoning, at least 3,500 years old - one of the oldest continuously inhabited settlements on earth, older than Athens, older than Jerusalem. It sits on the western bank of the Ganga in eastern Uttar Pradesh, 320 kilometres southeast of Lucknow, and its identity is layered in ways that resist the tidy categorisations that quick commerce operators prefer. It is simultaneously a pilgrimage city drawing five to eight million visitors a year, a university city anchored by the 1,300-acre Banaras Hindu University campus at Lanka, a textile city where an estimated 100,000 handloom weavers produce GI-tagged Banarasi silk, and - since 2014 - the parliamentary constituency of Prime Minister Narendra Modi, a fact that has channelled unprecedented infrastructure investment into a city that had been visibly underinvested for decades.
The infrastructure push is the first thing any market analyst should register. The Kashi Vishwanath Corridor, a Rs 800 crore redevelopment connecting the temple directly to the ghats, was completed in 2021. The Ring Road, long stalled, is now substantially complete, creating a bypass that decongests traffic from the GT Road axis. A metro rail DPR has been approved. The Varanasi-Prayagraj expressway shortens the drive to four hours. These projects have not transformed Varanasi into a modern grid city - the old city between Dashashwamedh Ghat and Manikarnika Ghat remains a labyrinth of gallis barely two metres wide, inaccessible to any motorised vehicle - but they have created pockets of navigable, apartment-dense, middle-class habitation along the GT Road, in Sigra, Cantonment, Nadesar, Bhelupur, and in the Lanka-Karaundi corridor south of BHU.
Those pockets are where quick commerce lives. Varanasi’s estimated 1.8 million population includes roughly 30,000 BHU students, several thousand coaching-institute students preparing for IIT and NEET examinations, a stable base of government employees, and a growing professional class in Sigra and the cantonment area. The city’s incomeIndex of 45 (on the QuickCommerceMap 0-100 scale) is among the lowest of any city with 20 or more dark stores - but income distribution matters more than the average. The student and professional pockets are the addressable market; the weaving community, the pilgrim economy, and the old-city bazaar population are, for now, structurally outside it.
Quick commerce story
Varanasi was a late entrant to quick commerce even by Uttar Pradesh standards. Lucknow had Blinkit stores by mid-2022. Noida and Ghaziabad, functioning as Delhi NCR satellites, had all three platforms operating at scale before Varanasi saw its first delivery. The city’s first dark stores appear to have opened in early 2024 - Blinkit leading, as it typically does in UP, with two to three stores near Lanka and Sigra. Swiggy Instamart followed in mid-2024, leveraging its existing food-delivery logistics base in the city. Zepto arrived last, opening stores branded VNS-BHU, VNS-Sarnath, and VNS-Naibasti - the three-letter city code plus locality naming convention that Zepto uses for every market it enters deliberately rather than experimentally.
As of the March 2026 snapshot, Varanasi has 21 dark stores: Blinkit leads with 11, Swiggy Instamart has 7, and Zepto trails with 3. The stores span a 12.6 km north-south corridor and a 7 km east-west band, but this overstates the actual coverage. Nine of the 21 stores cluster in or immediately around Sigra - the commercial and administrative heart of the modern city - and the Lanka-BHU corridor accounts for another five to six. The northernmost stores sit near Pandeypur and Sarnath; the southernmost is at Susuwahi, near the southern fringe of the old city. Nothing operates west of the GT Road industrial zone. Nothing operates east of the Ganga.
The concentration is not accidental. Sigra offers the combination that dark store operators need: wide enough roads for delivery riders, apartment and colony housing (Chandrika Colony, Sri Ram Nagar Colony), commercial activity that generates daytime orders, and proximity to institutional demand (Varanasi Nagar Nigam, district courts, hospitals). Lanka, at BHU’s main gate, offers something different but equally viable: a captive student population, dozens of PG accommodations and hostels, and a culture of app-based ordering that students bring from their home cities. Between these two anchors - Sigra and Lanka - Varanasi’s quick commerce market has a viable, if narrow, foundation.
What it does not yet have is breadth. Twenty-one stores serving a city of 1.8 million people yields a density of roughly 11.7 stores per million - below the national benchmark of 15 for a mature tier-2 market. More telling than the aggregate number is where the stores are not. The entire old city - the dense, commercially active, heavily populated zone between the ghats and Godowlia crossing - has zero dark stores. Ramnagar, across the Ganga, has zero. Sarnath, despite a Zepto store in its vicinity, is more accurately described as a fringe probe than a served market. The effective addressable population for Varanasi’s quick commerce network is probably 400,000-500,000, not 1.8 million.
Underserved areas
The old city is the most conspicuous gap, and the hardest to close. The zone between Dashashwamedh Ghat and Manikarnika Ghat - including Vishwanath Gali, Thatheri Bazaar, Chowk, and the maze of residential lanes behind the ghats - is home to an estimated 200,000-300,000 people living at densities exceeding 40,000 per square kilometre. These are among the highest population densities in any Indian city outside Dharavi. But the lanes are one to two metres wide. Auto-rickshaws cannot enter. Motorcycles squeeze through with difficulty. A delivery rider carrying a Blinkit or Zepto bag would need to navigate on foot from the nearest accessible road, adding five to ten minutes to every delivery - destroying the ten-minute promise that is the entire point of quick commerce.
Some operators may eventually experiment with foot-courier or bicycle-based last-mile delivery in the old city, but the economics are prohibitive at current order values. The old city’s retail needs are served by an extraordinarily dense kirana network - small shops embedded in every gali, restocked daily by headload porters - that has operated continuously for centuries and against which a Rs 200 grocery order delivered in thirty minutes by a walking courier cannot compete.
Trans-Ganga Ramnagar, home to the Ramnagar Fort and a growing residential population on the eastern bank, is a different kind of gap. The infrastructure exists - roads are wider, housing is more dispersed, and the Malviya Bridge and new Rajghat Bridge provide access. But the population density and income profile do not yet justify a dark store. Ramnagar is a future expansion target, not a current underserved market.
Sarnath, 10 km north, combines Buddhist pilgrimage tourism with a small residential population. The Zepto store near Asapur (branded VNS-Sarnath) appears to serve the Pandeypur-Sarnath corridor rather than Sarnath proper. The area’s demand is seasonal and tourist-driven - not the steady, repeat-order base that dark stores need.
Worker dimension
Varanasi’s 21 dark stores employ an estimated 168-315 workers - pickers, packers, scanning associates, shift incharges, and store managers. At the city’s tier-2 salary scale, entry-level pickers earn Rs 11,000-16,000 per month, shift incharges Rs 16,000-22,000, and store managers Rs 25,000-45,000. These wages are 20-30% below Mumbai or Bangalore equivalents but must be evaluated against Varanasi’s cost of living, which is among the lowest of any city in the QuickCommerceMap dataset.
A shared room in Lanka or Sigra costs Rs 2,000-4,000 per month. A basic meal at a local dhaba runs Rs 40-60. The purchasing power of a Rs 14,000 picker salary in Varanasi is roughly equivalent to Rs 20,000-22,000 in Mumbai. Labour availability is not a constraint - the city has a large population of young men with limited formal employment options, and the weaving industry’s decline has released workers who are physically capable and accustomed to warehouse-style shift work.
The constraint is attrition driven by aspiration. A worker who starts as a picker at a Blinkit store in Varanasi and proves capable will, within six to twelve months, receive offers from stores in Lucknow, Noida, or NCR - where the same role pays 30-50% more. Varanasi trains workers; larger cities absorb them. This is the tier-2 dark store labour paradox, and it applies here with particular force because the city’s educational infrastructure (BHU, coaching centres) produces a workforce that is literate, numerate, and upwardly mobile.
Consumer dimension
Varanasi’s incomeIndex of 45 and affordabilityIndex of 55 tell a clear story: this is not a city where quick commerce pricing aligns with mass-market purchasing power. The average Varanasi household’s grocery budget is oriented toward daily bazaar purchases - Rs 50-100 at a time, bought from a kirana or a gali vendor who offers informal credit, personalized service, and prices calibrated to local incomes. Quick commerce’s minimum order values (typically Rs 99-149), delivery fees, and platform pricing (5-15% above kirana for staples) create a structural mismatch for the majority of the population.
The addressable consumer base is narrow but real. BHU students are the most obvious segment: young, digitally native, accustomed to app-based ordering from their home cities (many come from tier-1 metros), and living in PGs or hostels where cooking is impractical and late-night snack runs are culturally normative. The Lanka-Naria-Karaundi corridor’s dark store density reflects this demand. Professional households in Sigra, Cantonment, and Bhelupur represent the second segment: dual-income families, government employees, and private-sector professionals whose time-value calculation favours quick commerce even at a price premium.
The third potential segment - tourist and pilgrim visitors - is large in volume but almost entirely unaddressable. Pilgrims buy puja items from ghat-side vendors, eat at ashram kitchens or street-food stalls, and shop for souvenirs in Vishwanath Gali. Their purchasing behaviour is experiential, not transactional. Until quick commerce develops a tourism-specific assortment (which no Indian operator has attempted), this population remains outside the addressable market.
Industry context
Among Uttar Pradesh’s quick commerce cities, Varanasi occupies a distinctive middle position. Lucknow, the state capital, has 94 stores and a mature multi-platform market. The NCR satellites - Noida (90), Ghaziabad (77), Greater Noida (36) - function as extensions of Delhi’s ecosystem. Kanpur, with its larger population (3.1 million), has more stores but a similar incomeIndex. Varanasi’s 21 stores place it ahead of Prayagraj, Gorakhpur, Bareilly, and Agra, but well behind the state’s leaders.
The more instructive comparison is with other religious and heritage cities nationally. Amritsar (1.5 million, incomeIndex 68) has quick commerce presence driven by Punjab’s higher per-capita income. Madurai (1.6 million, incomeIndex 73) has an Instamart-only probe with 5-10 stores. Haridwar and Ajmer, despite pilgrimage traffic, have no meaningful quick commerce presence. Varanasi, with 21 stores across all three platforms, is arguably the most developed quick commerce market among India’s major pilgrimage cities - a distinction that says as much about the category’s limits in religious cities as it does about Varanasi’s relative strength.
The BHU factor differentiates Varanasi from pure pilgrimage cities. Amritsar has Guru Nanak Dev University but at a fraction of BHU’s scale. Madurai has Madurai Kamaraj University but without the residential campus density. Varanasi’s 30,000-student residential university creates a year-round demand floor that pure pilgrimage cities lack. Strip out BHU, and Varanasi’s quick commerce viability probably drops to Prayagraj levels - a city where operators have barely entered despite 1.4 million people.
The growth trajectory from here depends on two factors. First, whether the Ring Road and expressway infrastructure translates into new residential development in areas like Pandeypur, Sarnath Road, and Shivpur - creating the apartment-dense housing stock that dark stores need. Second, whether Blinkit and Zepto commit to scaling beyond their initial footholds or treat Varanasi as a permanent 20-25 store market. The tier-2 expansion playbook suggests the former: once operators enter a city with 10+ stores, they tend to scale to 30-40 within 18 months if contribution margins are positive. Varanasi’s next twelve months will reveal which trajectory it follows.
Methodology
This report draws on the QuickCommerceMap verified dataset of 4,081 dark stores across India, last fetched from Blinkit, Zepto, and Swiggy Instamart public-facing APIs in March 2026. Varanasi’s 21 stores were individually reverse-geocoded using Ola Maps (primary), Mappls (fallback), and Nominatim (last resort) to obtain formatted addresses, localities, pin codes, and area assignments. Geographic spread was computed from coordinate data: the 21 stores span 12.6 km north-south (25.2533 to 25.3667 latitude) and 7.0 km east-west (82.9569 to 83.0203 longitude).
Platform arrival timeline estimates are derived from store-ID sequence analysis. Blinkit uses numeric IDs (Varanasi IDs fall in the 35,000-44,000 range), Swiggy Instamart uses numeric IDs (1.39-1.40 million range), and Zepto uses UUIDs that do not encode sequence but whose associated store names (VNS-BHU, VNS-Sarnath, VNS-Naibasti) confirm deliberate market entry. Demographic data derives from Census of India 2011, projected to 2026 using WorldPopulationReview methodology. Economic context uses MoSPI state-level NSDP figures, as city-level GDP data is not publicly available for Varanasi. Infrastructure references draw on UP government press releases and Smart City Mission documentation.
All indices (incomeIndex, smartphoneIndex, apartmentIndex, affordabilityIndex) are editorial judgements on a 0-100 scale, documented in the expansion enrichment panel. They are not derived from a single quantitative source but represent the research desk’s assessment informed by the sources listed above.