What to sell online in India

The four-filter matrix: margin, weight, fragility, repeat-purchase

Before you fall in love with a product, score it on four axes. This single habit prevents most first-year failures.

A product doesn't need to win on all four. But if it loses on three, walk away - the economics will grind you down.

• Margin: after cost of goods, packaging and payment fees, do you keep at least 40-50%? Low-margin items only work at high volume, which a new store won't have.
• Weight and size: courier charges are by weight and dimension. A ₹300 heavy item can carry ₹120 of shipping - brutal on a prepaid sale, worse on a returned COD parcel.
• Fragility: glass, ceramics, liquids and delicate electronics mean breakage, disputes and returns. Every fragile SKU needs bulletproof packaging that adds cost.
• Repeat-purchase: consumables (food, beauty, supplements) bring the same customer back. One-time buys (a wall clock) mean you pay to acquire a customer once and never again.

Check the competition honestly before you commit

Spend an evening doing a saturation check. Search your product on Amazon, Flipkart and Meesho. If page one is a wall of near-identical items at rock-bottom prices with thousands of reviews, you are entering a price war you can't win as a newcomer.

Saturation isn't automatically a no. It signals proven demand. The move is to find an angle the big listings ignore - a regional taste, a specific size range, a bundle, a story, faster local delivery, or a quality tier the discounters can't touch.

• Count the sellers and read their 2- and 3-star reviews - the complaints are your opening.
• Note the price floor. If you can't beat it and can't differentiate above it, pick something else.
• Check if the category is dominated by one giant brand versus many small sellers - fragmented categories are easier to enter.

Lean into your unfair advantage

The best product for you is one where you start ahead of a random competitor. Ask what you have that others must buy or build.

This is why generic dropshipping rarely works in India - everyone has the same supplier and the same price. Your edge is the moat.

• Sourcing access: a family unit in a manufacturing cluster (Tirupur knits, Moradabad brass, Firozabad glass, Jaipur block-print) buys cheaper than anyone reselling.
• Craft skill: you make it yourself - baked goods, handmade jewellery, art, tailoring. Hard to copy, commands a premium.
• Community: you already run a shop, a coaching class, a temple group, a WhatsApp community. That's warm demand you can sell to on day one.
• Local knowledge: a regional specialty (Bhagalpuri silk, Kashmiri saffron, Assam pickles) that outside sellers can't authentically offer.

Category snapshots for the Indian market

A quick reality check on the big categories, with the traps each one hides.

• Fashion and apparel: huge demand, but sizing drives returns and COD reversals. Win with a tight range and clear size charts rather than a giant catalogue.
• Food and packaged snacks: excellent repeat-purchase, but you need FSSAI registration/licence, shelf-life planning and food-grade packaging. Verify current FSSAI rules for your scale.
• Beauty and personal care: strong margins and repeat buys; cosmetics have their own labelling and regulatory norms - check before importing or private-labelling.
• Home and decor: good margins but weight and fragility hurt. Favour lighter, sturdier items or premium pieces that justify careful packing.
• Jewellery: high margin, low weight - a strong online fit. Imitation and silver are easier to start with; hallmarking rules apply to gold, so verify current BIS requirements.
• Regional specialties: lower competition and real authenticity, but demand can be seasonal and niche - plan for lumpy sales.

Validate before you buy inventory

The costliest mistake is buying a warehouse of stock for a product nobody wants. Prove demand first, cheaply.

Start with a small batch or a pre-order. Put up a real store, take orders, and only then scale sourcing. Season matters too - Diwali, weddings and school reopening create predictable spikes; build stock toward them, not against them.

• Run a pre-order or made-to-order model so buyers commit before you spend on bulk stock.
• Test with 10-20 units and watch the return rate and the questions buyers ask.
• Sell to your existing WhatsApp contacts and local network first - real money beats survey "maybes".

The unit-economics worksheet (a worked example)

Never judge a product by its selling price. Judge it by what's left after every real cost. Here's a printed cotton kurti sold at ₹899.

Cost of goods ₹350. Packaging ₹25. Payment-gateway fee on a prepaid order roughly 2% ≈ ₹18. Courier for a ~500g parcel, say ₹70 (Shiprocket-style rates vary by zone and weight). That's ₹463 in costs, leaving about ₹436 gross profit - a healthy ~48%.

Now the reality of COD and returns: if 1 in 5 parcels comes back, you eat two-way shipping (~₹140) plus repacking on that order, and it drags your average profit down meaningfully. Run the sum with an honest return rate, not a hopeful one. If the number still works, you have a business. If it only works at zero returns, you don't.

• Selling price minus (goods + packaging + payment fee + shipping) = gross profit per order.
• Subtract a return allowance (estimate 15-30% for fashion, lower for consumables).
• Whatever's left must also cover ads/marketing and your time - if it doesn't, reprice or re-pick.

Put it live and start selling

Once a product clears the matrix and the math, you need a place to actually take orders - with COD, UPI and cards, and shipping that works to real pincodes. That's where Kartrocket fits. You describe your business in plain language (Hinglish is fine) and a working store builds itself live; you point at anything and say what to change.

It's free to build and free to publish on yourname.kartrocket.ai with SSL included, so you can validate a product without upfront cost. Checkout, COD/UPI/cards and Shiprocket shipping across 24,000+ serviceable pincodes are built in, and prices are computed on the server so checkout can't be tampered with. You pay only real pass-through costs - courier charges, payment-gateway fees on prepaid orders, and your own domain if you connect one.