The best way to start an online business in India

The honest answer depends on one question

Where will your first hundred customers come from? Every other decision follows from that, and almost every bad decision comes from getting it backwards.

If the answer is people who already know you — your Instagram followers, your WhatsApp contacts, your neighbourhood, your existing clients — then you do not have a demand problem, you have an operations problem, and the right move is your own store. If the answer is strangers searching for a product and you have no audience at all, you have a demand problem, and a marketplace solves it faster than anything you can build. Most people reading this are in the first group and behave as though they are in the second.

The four routes below are ranked for that first group, because it is the common case in India. If you are genuinely in the second, read the marketplace entry first — it is ranked second here, not because it is worse in absolute terms, but because it is worse for someone who already has attention and would be paying a commission for demand they do not need.

The four routes, ranked for the common case

1. 1

   Your own store, promoted from where you already are

   Best for most first-time sellers

   Keep finding customers on Instagram, WhatsApp and word of mouth, but send them to a store that takes the order properly: prices, stock, a real checkout, an order ID and a delivery address collected once. You carry the whole burden of demand — nobody is going to browse your store by accident — but you pay no commission, you keep the customer list, and your operations stop being a chat thread. This is the route most Indian sellers should start on because the cost of being wrong is close to zero: publishing a store on SitesPlaced is free, with unlimited products, Razorpay, cash on delivery, WhatsApp checkout and Shiprocket included.

   ₹0 to publish on SitesPlaced · 0% commission · payment-gateway fee appliesBest for: Anyone already getting DMs who is losing orders in chat.
2. 2

   A marketplace (Amazon, Flipkart, Meesho and the category ones)

   Best when you need demand more than margin

   The one route that brings you customers on day one. Millions of people are already searching there, the payment and delivery infrastructure exists, and a first-time seller does not have to explain why a stranger should trust them. The price is steep in three currencies: a per-sale commission that can take a large slice of a thin margin, a search results page where the lever is price, and no ownership of the customer — you cannot email or message a buyer who bought from you. Selling through a marketplace can also change your registration position, so check the GST rules for e-commerce operators before you list.

   Free to list; category commission plus fulfilment, closing and payment fees on every saleBest for: Commodity products with room in the margin, or anyone who needs volume to test demand fast.
3. 3

   Instagram and WhatsApp alone, with no store

   Fastest start, lowest ceiling

   This is how an enormous number of Indian businesses genuinely begin, and there is nothing wrong with it for the first month. Post the product, take the DM, send a UPI ID, ship it. It costs nothing and needs no setup. It stops working at roughly the point it starts working: orders and addresses live in chat threads, payment confirmation is a screenshot you have to trust, stock lives in your head, and there is no way to answer the question of what you actually sold last month. Treat it as a starting point, not a destination.

   ₹0 — you pay in time and lost ordersBest for: Testing whether anyone wants the product at all, before spending anything.
4. 4

   Reselling or dropshipping someone else's catalogue

   Lowest capital, hardest economics

   You sell products you do not hold, and a supplier or aggregator ships them. It removes the largest cost of starting — inventory — and the largest risk, which is unsold stock. What it does not remove is competition: you are selling the same catalogue, at similar prices, as everyone else with the same supplier, so the margin is thin and the differentiation has to come entirely from your audience or your service. Quality control and delivery timelines are also someone else's decisions, and returns land on you either way.

   Little or no stock investment; margins are set by your supplierBest for: Sellers with an existing audience and no capital, who can accept thin margins.

The routes compared

The recommendation, stated plainly

For the majority of people asking how to start an online business in India: start selling on Instagram or WhatsApp today, publish a free store this week, and only consider a marketplace once you know which products sell and what margin they carry. That order costs you almost nothing, and it means the audience you build belongs to you rather than to a platform that can change its rate card.

Online business ideas that hold up in India in 2026

The ideas that survive share a shape rather than a category: low or no inventory, a buyer you can already reach, and a margin that survives a courier charge. These are the shapes that keep working.

• ✓Made-to-order or small-batch physical products — jewellery, apparel, home and gifting. Inventory risk is low and the story is the differentiator.
• ✓Food, where you can meet the FSSAI requirements for your scale. Repeat purchase rates are the highest of any category and the buyers are local, which keeps shipping cheap.
• ✓Services and consulting — design, tutoring, accounting, fitness, photography. Zero inventory, and a website exists mainly to answer the trust question.
• ✓Digital products and courses. No shipping, no returns, and margin that survives being sold to one person at a time.
• ✓Local services with an online booking or enquiry front, where the internet is the shopfront and the delivery is physical and nearby.
• ✓Resale, but only where you have a real audience advantage — otherwise you are competing with everyone who has the same supplier.

If you want the legal and registration side of this — structure, Udyam, PAN, GST, licences — that is a separate job and we wrote it up separately in how to start a business in India.

Launching in a week, realistically

1. 1

   Pick the ten products you would actually defend

   Not the whole catalogue. Ten products you can photograph well, price with a real margin and ship reliably. A narrow store that ships on time beats a broad one that does not.

2. 2

   Price with shipping, packaging and gateway fees inside the number

   Work out the landed cost per order before you set a price, including the courier charge, the packaging, the payment-gateway percentage and an allowance for returns. This is where most first-year Indian ecommerce margin disappears.

3. 3

   Publish a store and connect payments

   Get the store live with UPI and cards through your own Razorpay account, cash on delivery if your category expects it, and WhatsApp checkout for buyers who want to ask before they pay. Publishing is free on SitesPlaced and there is no product cap.

4. 4

   Write the delivery and return policy before anyone asks

   Two short paragraphs stating dispatch time, delivery time and what happens if something is wrong. This removes the single largest objection a first-time buyer in India has.

5. 5

   Point every channel you already have at one link

   Instagram bio, WhatsApp status, your status message, the local groups, your existing customers. One link, one place orders land, one record of what sold.

6. 6

   Ship the first ten orders yourself and ask every buyer why they bought

   The first ten orders are research. What they tell you about price, objections and which product should have been on the homepage is worth more than another week of design.

What each route actually costs

The software is rarely the expensive part any more, which is why comparing monthly plan prices is the wrong way to make this decision. The costs that decide whether an Indian online business works are the courier charge per order, the packaging, the payment-gateway percentage, the returns, and — on a marketplace — the commission.

• ✓Marketplace: no upfront cost, but a category commission plus fulfilment, closing and payment fees on every order. Model this at your real average order value before you list, not at your best-case one.
• ✓Your own store: on SitesPlaced, ₹0 to publish with unlimited products and 0% commission. Your Razorpay account charges its standard gateway fee on online payments. The ₹499/month Ecommerce plan — an early-bird price for the first 30 stores, ₹999 after — adds a custom domain, badge removal, AI copy, store analytics, Meta Pixel with the Conversions API, and WhatsApp Business API order alerts from your own number.
• ✓Instagram and WhatsApp only: free in rupees, expensive in hours, and the cost shows up as orders you lost track of rather than as an invoice.
• ✓Reselling: little capital, but the supplier sets the margin and you inherit their delivery times and quality.

Before you commit to any of them, run your actual numbers through the free profit margin calculator and, if you are quoting tax-inclusive prices, the GST calculator. A route that looks cheap at list price often is not once a commission and a return sit on top of it.

Mistakes that decide the outcome early

• ✓Choosing a route by monthly plan price instead of by cost per order, including commission, shipping and returns.
• ✓Building only on a platform you do not control, and discovering in year two that you have no customer list.
• ✓Launching a hundred products instead of ten, so nothing is photographed well and nothing ships on time.
• ✓Treating a marketplace listing as a brand. It is a search result — the moment you stop paying for placement or dropping price, it stops.
• ✓Not checking whether your route changes your GST position. Selling through an e-commerce operator has historically been treated differently from selling on your own store — verify before you list.
• ✓Waiting for a perfect store. A live store with ten honest listings outperforms a beautiful one nobody has seen.