Amazon vs Flipkart vs Myntra for sellers in India: 2026 decision guide

There is no universal winner between Amazon, Flipkart and Myntra. The right first marketplace is the one that accepts the product, shows credible demand for comparable items and leaves positive contribution after the seller's actual costs.

This guide is a decision framework for Indian sellers. It does not rank platforms by unverified audience quality, return rate or sales potential, and it does not treat a zero-commission headline as proof that an order will be profitable.

Amazon vs Flipkart vs Myntra at a glance

Use this table to decide what to verify first. It is not a forecast of sales.

| Decision factor | Amazon India | Flipkart | Myntra | | --- | --- | --- | --- | | Broad fit | Wide product selection | Wide product selection | Fashion and beauty focus with defined selection criteria | | First eligibility check | Seller registration, category and product requirements | Seller registration, category and product requirements | Brand qualification and category selection | | Public fee evidence | Detailed category, price-band and fulfilment tables | Current fees-and-commission page plus seller-account terms | Seller agreement and account-specific commercial terms | | Fulfilment decision | FBA, Easy Ship or Self-Ship | Confirm the available fulfilment route and rate card for the account | Confirm the approved operating and fulfilment model during onboarding | | Catalogue emphasis | Accurate identifiers, attributes, images and offer data | Accurate catalogue, package and dispatch data | Brand, imagery, size/fit and fashion-or-beauty attributes where applicable | | Best reason to test first | The SKU passes eligibility, demand and contribution checks | The SKU passes eligibility, demand and contribution checks | The brand qualifies and the assortment is ready for a curated catalogue |

The table deliberately avoids generic labels such as “best for beginners” or “lowest cost.” Those claims can hide the category, selling price, fulfilment route and return behaviour that decide the result for an individual SKU.

Start with eligibility and category fit

Amazon and Flipkart accept sellers across broad product ranges, but registration does not automatically make every product eligible. Category approval, product compliance, brand documentation and listing requirements can still apply. Check the current seller workflow for the exact legal entity, category and product before building a launch plan.

Flipkart's official registration guide, dated 29 June 2026, says account creation is free and commission is charged when a sale occurs. That describes registration; it does not establish the full cost of selling a particular SKU.

Myntra requires a different first question: does the brand fit its selection criteria? The official Myntra partner portal says that a qualifying brand can be onboarded within 15 business days, with longer timelines possible during major sale events. Treat that as an onboarding statement for qualifying brands, not an acceptance or completion guarantee.

Before comparing platforms, create an eligibility file containing:

• legal-entity, GST, bank and address records required by the platform;
• brand ownership or authorisation evidence;
• product identifiers and category-specific compliance records;
• verified dimensions, weight, package contents and tax data;
• image, attribute, size and fit assets required by the intended category;
• the dated source or Seller Central screen used for every requirement.

If the SKU cannot pass the eligibility check, marketplace size and fee headlines are irrelevant.

Compare total order economics, not commission alone

Marketplace fee structures change. Rebuild the comparison from current official terms on the day the price or fulfilment model changes.

Amazon India's fees and pricing page shows category- and price-dependent referral fees, closing fees that vary by price and fulfilment channel, and weight-handling charges. It also states that updated closing fees became effective on 7 September 2026. Amazon's public calculator is useful for planning, but Amazon labels its results as estimates; actual transaction and settlement data should replace the estimate after launch.

Flipkart's current fees-and-commission page advertises 0% commission for products priced below ₹1,000 and across fashion products. Commission is only one line in the order economics. Verify shipping, collection, fixed or other applicable charges, taxes, adjustments and return-related costs in the current rate card and seller account before deciding that a SKU is profitable.

For Myntra, use the commercial agreement, account rate card and transaction reports supplied for the approved seller and category. Do not borrow an Amazon or Flipkart fee assumption to fill a missing Myntra value.

Build the same contribution model for every platform:

Net contribution per delivered order = realised selling price − product cost − marketplace charges − fulfilment and packaging − advertising allocation − return and damage allowance

Use realised selling price after seller-funded promotions. Keep taxes and recoverable credits on a consistent basis agreed with a qualified adviser. Separate resellable returned inventory from actual write-offs, and replace every allowance with observed cohort data once enough orders have resolved.

Illustrative comparison only

Suppose one SKU is tested on two platforms. Platform A has a lower commission line, while Platform B has a higher commission line but lower fulfilment and return cost for that seller's resolved cohort. Platform B can still produce the stronger contribution.

No rupee values are assigned here because the correct amounts must come from the seller's current rate card, package data and transaction reports. The lesson is structural: compare the complete delivered-order economics, not one fee label.

Match the fulfilment model to the operation

Amazon's official shipping and fulfilment overview describes FBA, Easy Ship and Self-Ship. Each route changes who stores, packs and delivers the order, and it changes the fee and inventory model that should be evaluated.

For Flipkart and Myntra, confirm the fulfilment options actually available to the seller account and category. Record who owns:

• inventory allocation and replenishment;
• packaging and label compliance;
• dispatch cut-offs and daily exception handling;
• cancellation and return evidence;
• damaged or wrong-item claim files;
• settlement matching and unresolved adjustments.

A platform is not operationally suitable merely because the account can be opened. The team must be able to keep stock, catalogue facts, dispatch promises and evidence records accurate without weakening another sales channel.

Check demand without inventing market size

Public marketplace search results can support a launch decision, but they do not provide reliable keyword volume or a sales forecast. Review the same exact product type on each eligible marketplace and date the observations.

Capture:

• close substitutes in the first useful result set;
• visible selling-price and pack-size ranges;
• review recency, not only lifetime review totals;
• repeated buyer questions and visible specification gaps;
• sponsored-result density;
• stock availability and delivery promises for representative serviceable pin codes;
• whether one established listing dominates or demand appears distributed.

Keep screenshots and distinguish sponsored placements from organic results. A search page can show how products are presented today; it cannot reveal a competitor's profit, return rate or future demand.

Score catalogue readiness before uploading

The first marketplace should be the one the team can launch accurately. Score each candidate SKU from 0 to 2 on the following checks:

1. Product record: identifiers, brand, variation and included contents are verified.
2. Images: the main, detail and lifestyle assets match the exact variation.
3. Attributes: material, size, colour, compatibility and package data are complete where relevant.
4. Compliance: required category and product records are available and current.
5. Operations: stock, pricing, dispatch and returns have named owners.

For Myntra, add assortment depth, size and fit information, and the brand-selection materials requested for the category. For Amazon and Flipkart, verify category-specific templates and returned processing errors rather than assuming a successful file upload means every offer is buyable.

Run one controlled SKU-cohort test

Do not upload the full catalogue merely to make the account look complete. Start with a small, representative cohort of in-stock SKUs whose costs and package data are understood.

Define before launch:

• the exact SKUs and stock ceiling;
• price and promotion boundaries;
• the fulfilment route;
• the advertising cap and contribution guardrail;
• catalogue acceptance checks;
• owners for dispatch, returns and settlement review;
• stop, repair and expansion rules.

Measure impressions, clicks, conversion, cancellations, returns, advertising spend, confirmed settlement and contribution together. A short test can reveal catalogue or fulfilment defects, but it cannot establish a stable return rate until the relevant orders have resolved.

Use a dated marketplace scorecard

Score Amazon, Flipkart and Myntra from 0 to 2 on six dimensions. Attach a source, owner and evidence date to every score.

| Dimension | 0 | 1 | 2 | | --- | --- | --- | --- | | Eligibility | Blocked or unknown | Requirements incomplete | Confirmed for test cohort | | Demand evidence | No credible comparable set | Mixed or uncertain | Active comparable set observed | | Contribution | Negative or unknown | Sensitive to assumptions | Positive under documented inputs | | Catalogue readiness | Major data gaps | Repairable gaps | Test files ready | | Fulfilment capacity | No clear owner | Capacity constrained | Named owner and process | | Reconciliation readiness | No order-to-payout method | Partial tracking | Item-level ledger ready |

The highest evidence-backed score identifies the safest first test, not a guaranteed winner. Break a tie by choosing the smaller reversible test or the platform with fewer unresolved operational dependencies.

Add the second marketplace only after the first is understood

Expand after the first platform produces a repeatable operating record:

• selected SKUs remain in stock reliably;
• catalogue errors and dispatch exceptions are controlled;
• actual settlement lines reconcile to order-level records;
• returns are classified by reason and confirmed cost;
• advertising stays within the documented contribution guardrail;
• platform-specific inventory, pricing and content have clear owners.

Do not copy a listing mechanically. Preserve verified product facts, then adapt images, attributes and searchable wording to the destination platform's current requirements. Keep separate performance and settlement ledgers so one channel cannot hide another channel's losses.

Final decision checklist

Choose the first marketplace only when all seven answers are documented:

1. Is the brand, legal entity and product eligible?
2. Do current search results show a credible comparable-product set?
3. Does the SKU remain contribution-positive under the current rate card and fulfilment model?
4. Are catalogue and compliance assets ready for the exact category?
5. Can the team meet dispatch, stock and return workflows consistently?
6. Is there a small SKU cohort with explicit stop and expansion rules?
7. Can the first complete order, return and settlement cycle be reconciled at item level?

Recalculate the scorecard after the first resolved cohort. The best first marketplace is the one that creates a defensible operating model for the seller's actual products; the best long-term channel mix may be different once evidence exists across more than one platform.