Consumer, Retail and E-commerce in India
India retail FDI is gated by format, not sector: single-brand, multi-brand, marketplace vs inventory, food. The routes, caps and entry path explained.
How foreign investment into India's consumer market is permitted - and why the route depends on the retail format you choose, not on the sector you are in.
India is the demand story most consumer companies want exposure to, and the capital is flowing: consumer and retail was the single largest sector by deal volume in India through 2025 (Grant Thornton Bharat). But the foreign-direct-investment rules do not treat "retail" as one thing. They gate capital by the format you trade in - how you sell, not what you sell. The same foreign rupee is 100% automatic in one format, allowed only to 51% on the government route in another, and barred outright in a third. The decoder for this entire page is therefore a sequence: decide the format first, and the FDI route follows from it.
The single most counter-intuitive consequence - and the one that catches the most well-advised entrants - is this: an FDI-funded entity generally cannot own inventory and sell it directly to Indian consumers online. Owning your own stock and selling it business-to-consumer on your own website is the inventory-based e-commerce model, and that model is closed to foreign investment. The operational test is a single question - who owns the stock at the point of sale to the consumer? If the FDI-funded entity owns it and sells B2C, that is the barred inventory model (unless the India-made-food carve-out applies); if independent sellers own it and the platform only connects them, that is a permitted marketplace. What an FDI-funded company can do online is run a marketplace that connects independent sellers, manufacture in India, or sell its own single brand through an owned-brand entity. The cheque size is irrelevant to which of those applies. So the first question is which entrant you are: a global brand opening its own stores and site; a grocery or hypermarket operator wanting multi-brand shelf space; an online-first platform; a food or FMCG manufacturer building a make-in-India base; or a D2C brand deciding how to reach Indian consumers at all - each maps to a different box below.
At a glance
- FDI is gated by retail FORMAT, not sector. Pick the format first; the route - automatic, government, or barred - follows.
- Single-brand retail: 100% automatic. Once foreign holding crosses 51%, a 30% domestic-sourcing condition applies - materially eased by the 2019 relaxations.
- Multi-brand retail: 51% on the government route - law since 2012, but heavily conditioned (back-end investment, small-industry sourcing, city limits) and the state has to opt in. Constrained and little-used, not unavailable.
- E-commerce - the line that matters: the marketplace model is 100% automatic; the inventory-based (own-stock B2C) model is barred for FDI. Press Note 2 (2018) polices the boundary.
- Food-product retail/e-commerce: 100% on the government route, limited to food made or produced in India - the one narrow case where an FDI-funded entity can own and sell inventory to consumers online.
- Manufacturing and B2B / cash-and-carry wholesale: 100% automatic - the cleanest box for a D2C or FMCG operator.
- Two policy flags on every deal: Press Note 3 (2020) beneficial-owner look-through - now eased by Press Note 2 of 2026 for non-controlling stakes up to 10% (screen UAE-routed capital with upstream land-border ownership) - and the fact that a private limited company, not an LLP, is the vehicle for the conditioned formats.
Why India, and why now
India is the demand-side bet: a large, formalising consumer base, rising organised-retail penetration, and a digital-commerce surface that now reaches well beyond the metros. The capital signal is concrete - consumer and retail ranked as the number-one sector by deal volume in India through 2025, on Grant Thornton Bharat's deal tracking, ahead of the technology and pharma sectors that usually lead the count. The market behind that signal is large and fast: India's e-commerce sector is on track for around US$160 billion by 2026, with a direct-to-consumer layer near US$90 billion and a quick-commerce layer growing 70-80% a year off a US$7-8 billion base. Inbound strategic interest in 2025 spanned packaged foods and staples as well as branded consumer platforms, and private-equity activity into the sector rose year on year. One distinction matters before the structuring starts: the demand surface and the FDI permission are different things. ONDC, quick commerce and the D2C boom expand how Indian consumers buy; they do not create new FDI routes or relax the inventory-model rule. The pattern is consistent: foreign capital wants the consumption growth and is finding formats - manufacturing, single-brand, marketplace, food - through which the FDI rules permit it. The constraint is never appetite; it is structuring.
Which format are you? The FDI maze
Almost every mistake on an India consumer entry is a format mistake: an investor designs the business it wants, then discovers the FDI route that format carries. Reverse it. Establish the format, read the route off it, and only then structure.
The reading: sell one brand and you are largely free but tied to a sourcing condition; sell many brands physically and you are on a narrow, heavily-conditioned government route that few have used; sell online and the only question that matters is whether you own the stock - own it and you are barred, connect others' stock and you are automatic; sell India-made food and a special door opens; and if you only make or wholesale, you have the cleanest entry of all.
| Format | Typical investor | Key legal issue (the deal-breaker) |
|---|---|---|
| Single-brand retail (SBRT) | Global brand selling one label through owned stores and its own site | 100% automatic; 30% of value sourced from India once foreign holding exceeds 51%; must be a single brand, sold under one name globally |
| Multi-brand retail (MBRT) | Global grocery, hypermarket or department-store operator | 51% on the government route - law since 2012 but heavily conditioned (minimum investment, back-end, small-industry sourcing, city limits) and the state must opt in; little-used, not unavailable |
| E-commerce - marketplace model | Platform connecting independent third-party sellers | 100% automatic, but the Press Note 2 (2018) conditions: no single vendor may dominate, the platform may not control inventory or influence price, services must be arm's-length, no mandated exclusivity |
| E-commerce - inventory-based model (B2C) | Online-first business wanting to own the stock it sells to consumers | FDI barred. The route is to manufacture/wholesale upstream and reach consumers via a marketplace, or to sell an owned brand through an SBRT entity |
| Food-product retail / e-commerce | Food and FMCG strategics; food-first D2C brands | 100% on the government route, limited to food made or produced in India - permits inventory ownership and direct online sale that general e-commerce cannot |
| Manufacturing + B2B / cash-and-carry wholesale | D2C and FMCG operators building an India base | 100% automatic (the cleanest box); wholesale cannot sell B2C - the consumer-facing front end needs its own SBRT entity or independent sellers |
The 30% sourcing rule - and why it is more manageable than it reads
For single-brand retail, the headline condition scares entrants more than it should. Once foreign holding crosses 51%, the entity must source, on average, 30% of the value of its goods from India. Read cold, that sounds like a margin-killer for a brand that manufactures abroad. The 2019 relaxations changed the arithmetic: all India procurement counts - whether the goods are sold in India or exported (the earlier five-year cap on counting exports was removed); sourcing for the brand's global operations counts toward the requirement, with an incremental-sourcing set-off applied in the early years, so a brand that already buys components or product out of India for its worldwide business gets credit for it; sourcing from units in Special Economic Zones counts; and online sales are permitted ahead of opening a physical store (with a brick-and-mortar store to follow within two years), removing the old "bricks before clicks" sequencing. The condition does not disappear, and it has to be planned and evidenced. But for most global brands with any existing India supply relationship, the practical question shifts from "can I survive 30%?" to "how do I document the sourcing I am already doing?" That is a structuring exercise, not a deal-breaker.
The marketplace tightrope - Press Note 2 (2018)
The marketplace model is 100% automatic, but it lives inside a rulebook designed to stop a marketplace from quietly becoming an inventory retailer. Press Note 2 (2018, effective 1 February 2019) sets the conditions a foreign-funded marketplace must keep to. No vendor concentration: a seller is deemed to be controlled by the marketplace if it buys more than 25% of its purchases from the marketplace or its group entities - cross that line and the platform is reclassified as inventory-based, which is barred. No inventory control: the marketplace may provide the technology platform and logistics, but it cannot own or control the inventory sold on it. Arm's-length services: services the platform provides to sellers - fulfilment, advertising, financing, support - must be fair, non-discriminatory and at arm's length, and cannot be used to favour particular sellers. No mandated exclusivity: a marketplace cannot require a seller to sell exclusively on its platform. No price influence: the marketplace cannot directly or indirectly influence the sale price of goods, which has to sit with the independent seller. These conditions are enforced, and enforcement has tended to focus on exactly the boundary cases - equity or control links between a marketplace and its largest sellers, preferential treatment, and pricing influence that makes a "marketplace" look like a captive store. The compliance posture that matters is structural: keep genuine separation between the platform and the sellers on it, and be able to prove the five conditions are met in substance, not just on paper.
The food carve-out - the one clean inventory door
There is a single, narrow exception to the inventory bar. Retail and e-commerce of food products is permitted at 100%, on the government route, where the food is manufactured or produced in India. This is the only structure in which an FDI-funded entity can own inventory and sell it directly to Indian consumers - including online - rather than merely operating a marketplace. The carve-out is deliberately bounded to India-made food, and it runs on the government (approval) route, so it is not a general workaround for the inventory bar. But for a food or FMCG investor whose product is, or can be, made in India, it converts an otherwise-closed B2C online model into a permitted one. FSSAI compliance sits on top of the FDI position for any food business.
Which retail entry route fits your business model?
The format table reads off the law; this reads off the business. Most entrants are one of these, and the test in each case is different:
- Global single-brand owner: test the SBRT route, the 30% sourcing trail, online-ahead-of-store, and the India sourcing or manufacturing you already do.
- Multi-brand retailer: test the MBRT route, the state opt-in, the back-end-investment and sourcing conditions, and whether the commercial model is worth the approval burden at all.
- Online marketplace: test seller independence, the no-inventory-control and 25%-vendor-concentration lines, related-party sales and the Press Note 2 fairness conditions.
- D2C / FMCG manufacturer: test manufacturing or B2B wholesale first (the cleanest box), then the consumer-facing channel separately.
- Food-products business: test whether the product is food made or produced in India, and whether the government-route food carve-out is usable.
- UAE distributor / trading house: test wholesale, franchise, a brand licence or a local distributor before committing to any stock-owning consumer model - and run the Press Note 3 ownership screen on the capital.
How a foreign company enters
The vehicle for a consumer, retail or e-commerce entry is a private limited company. An LLP is not clean for any of the conditioned formats - FDI into an LLP is workable only where the activity is 100% automatic with no conditions, which describes unconditioned manufacturing but not single-brand, multi-brand or food retail, all of which carry conditions; a liaison or branch office cannot conduct B2C retail. So the practical entry is to incorporate an Indian private limited company, capitalise it through the correct FDI route for the chosen format, and complete the FEMA reporting; the mechanics of incorporation, capitalisation and reporting are covered on the structuring spokes and not restated here. For a D2C or FMCG entrant who wants the cleanest path, a common pattern is two entities working together: a 100%-automatic India manufacturing company that makes the product, and - where the brand wants its own consumer-facing storefront - a parallel single-brand retail company for that owned brand; direct online B2C is then reached through the manufacturing or wholesale layer selling to independent marketplace sellers, not by the foreign-funded entity owning and selling its own online stock. This is compliance architecture, not a workaround: the foreign-funded entity never owns the consumer-facing stock - genuine independent sellers do. Overlaying all of this is Press Note 3 (2020): it requires prior government approval for FDI coming from, or beneficially owned by, an entity of a land-border country, regardless of sector or route. Press Note 2 of 2026 (issued 15 March 2026) eased this at the margin - a land-border beneficial owner holding up to 10% with no control may now invest through the automatic route on a reporting basis, with mandatory DPIIT reporting; above 10%, or where control is exercised, prior approval continues. The look-through is the point: routing capital through an intermediate holding company does not cure it. For India-UAE deals specifically, the UAE itself shares no land border with India, but capital that flows via the UAE while being beneficially owned upstream in, for example, mainland China or Hong Kong falls within the regime and needs screening. Press Note 3 depth and the FEMA mechanics live on the FEMA-advisory spoke.
Legal workstreams for a consumer, retail or e-commerce entry
A clean entry resolves these, in roughly this order:
- format determination - fix the retail format and read the FDI route, cap and conditions off it, before incorporating anything;
- Press Note 3 screen - map the full beneficial-ownership chain for any land-border exposure, including capital routed through the UAE or elsewhere; decide whether prior approval is required or whether the Press Note 2 of 2026 sub-10%-no-control automatic carve-out applies;
- entity and capitalisation - incorporate the private limited company; capitalise via the correct route (automatic vs government); complete FC-GPR / FC-TRS reporting under FEMA;
- condition compliance - build the evidence framework for the conditions that attach to the chosen format: the 30% sourcing trail for SBRT, the Press Note 2 separations for a marketplace, the made-in-India and approval position for food retail;
- brand, IP and intra-group terms - the brand-licence, royalty and intra-group service arrangements on arm's-length terms, which drive both the tax position and, for marketplaces, the Press Note 2 fairness test;
- consumer-law, data and sector compliance - the Consumer Protection (E-Commerce) Rules 2020 for online sellers and marketplaces, the Digital Personal Data Protection Act for the consumer data the business collects, FSSAI for food, and labelling and packaging rules for goods (a set of 2021 draft amendments to the e-commerce rules remains proposed and not in force);
- repatriation and exit planning - how dividends, royalties and eventual exit proceeds repatriate, modelled on the structuring spokes.
The India-UAE corridor
For investors structuring an India entry through the UAE, the consumer sector carries one specific watch-item beyond the usual holding-company logic: Press Note 3 beneficial-ownership screening. A UAE holding entity is unremarkable in itself, but if the capital is beneficially owned further upstream in a land-border country, the UAE layer does not cure the requirement - the look-through reaches through it (subject now to the Press Note 2 of 2026 carve-out for non-controlling stakes up to 10%). Treat the corridor as a legitimate and common route, but diligence the ownership chain end to end. CEPA tariff treatment and the broader corridor mechanics are covered on the corridor and CEPA pages and not duplicated here.
Where this goes wrong
- Designing the business, then discovering the route. Building the model first and checking the FDI position second is the classic error - the format dictates the route, so determine it first.
- Assuming you can sell your own inventory online. The most common surprise: an FDI-funded entity generally cannot own stock and sell it B2C online. Plan for marketplace, owned-brand SBRT, or the food carve-out from the start.
- Treating the 30% sourcing rule as fatal - or ignoring it. It is neither: it applies above 51% holding, it is materially eased by the 2019 relaxations, and it must still be planned and evidenced.
- Running a marketplace too close to its sellers. Equity links, preferential services, exclusivity or price influence can reclassify a marketplace as inventory-based and breach Press Note 2 - enforcement concentrates on exactly these boundary cases.
- Banking on multi-brand retail. It is available at 51% but hemmed by conditions and a state opt-in, and few have used it - confirm the position in the relevant state before building a thesis on it.
- Missing the Press Note 3 look-through. Routing through the UAE or another third country does not escape beneficial-owner screening; an upstream land-border owner with more than 10%, or with control, still needs approval.
- Choosing an LLP for a conditioned format, and treating the still-draft 2021 e-commerce amendments as law when they are not notified. The Press Note 3 easing, by contrast, is now law (Press Note 2 of 2026) - apply it correctly rather than over- or under-reading it.
How ATB Corporate helps
ATB advises foreign investors on India consumer, retail and e-commerce entries from the first question that decides everything - which format you are in, and therefore which FDI route, cap and conditions apply. We map the beneficial-ownership chain for Press Note 3 (and the Press Note 2 of 2026 carve-out), set up the private limited vehicle and capitalisation on the correct route, build the compliance evidence for the conditions that attach to your format (single-brand sourcing, marketplace separations, the food carve-out), and coordinate the deal and diligence workstreams for an inbound transaction. Incorporation, FEMA reporting, structuring, tax and on-ground distribution are delivered through our dedicated India spokes, linked below.
Consumer, Retail & E-commerce — Answered
FDI in Indian retail is set by format: single-brand retail is 100% automatic; multi-brand retail is 51% on the government route with conditions and a state opt-in; e-commerce marketplaces are 100% automatic; inventory-based (own-stock B2C) e-commerce is barred for FDI; and retail or e-commerce of India-made food is 100% on the government route. There is no single "retail FDI" figure - the format sets the route, so the limit depends entirely on the format.
No in almost all cases: owning the stock you sell B2C online is the inventory-based model, which is closed to FDI. A foreign-funded entity can instead run a marketplace connecting independent sellers, sell its own single brand through an SBRT entity, or - in the one narrow exception - sell India-made food under the food-retail route.
Once foreign holding in a single-brand retail entity exceeds 51%, the entity must source, on average, 30% of the value of its goods from India. The 2019 relaxations make it manageable: all India procurement counts whether sold in India or exported, sourcing for the brand's global operations counts with an early-years set-off, and SEZ sourcing counts.
Yes, in law - up to 51% on the government route, in place since 2012. In practice it carries minimum-investment, back-end and small-industry-sourcing conditions and city restrictions, the state must opt in, and few investors have used it; it is constrained and commercially difficult, not legally unavailable. Confirm the position in the specific state before relying on it.
Press Note 2 (2018) sets them: no single vendor may buy more than 25% of its purchases from the marketplace or its group; the platform cannot own or control inventory; services to sellers must be arm's-length and non-discriminatory; no mandated exclusivity; and the platform cannot influence sale prices. Breaching these can reclassify the marketplace as inventory-based, which is barred.
The UAE shares no land border with India, so UAE capital is not caught merely for being UAE capital. But Press Note 3 works on beneficial ownership and looks through intermediate holding companies - so capital routed via the UAE while beneficially owned upstream in a land-border country (for example China or Hong Kong) needs screening. Press Note 2 of 2026 now lets a non-controlling stake of up to 10% proceed on the automatic route with reporting; above that, or with control, prior approval applies.
Manufacturing is the cleanest box - 100% automatic with no format trap. A common pattern is an India manufacturing company plus, where the brand wants its own storefront, a parallel single-brand retail entity, reaching the open online market through independent marketplace sellers rather than by owning and selling its own online inventory.
Not cleanly for the conditioned formats. FDI into an LLP is workable only where the activity is 100% automatic without conditions, which fits unconditioned manufacturing but not single-brand, multi-brand or food retail. The private limited company is the standard vehicle.
The Consumer Protection (E-Commerce) Rules 2020 are in force and apply to online marketplaces and sellers. A set of 2021 draft amendments - covering flash sales and fall-back liability, among other things - remains proposed and has not been notified; it should not be treated as current law.
No. They change the demand surface and the channels through which consumers buy, but they do not create new FDI routes - a participant is still either a marketplace (automatic) or an inventory-based seller (barred for FDI) under the same rules.
In Indian retail, the format sets the FDI route and the stock-and-seller model decides what is permitted, with Press Note 3 looking through any intermediate holding — best settled before brand and channel are locked.
Licensing, approvals and any tax treatment are decided by the authorities on the facts. Talk to our team when you are ready.
Get in touch