Performance

Short answer: Marketplaces such as Amazon and Flipkart give Indian D2C brands ready demand and trusted delivery, in return for fees, limited customer data and price pressure. Your own website costs more to bring traffic to, but you keep the customer relationship. Most brands need both, and should set the mix by profit per order on each channel, not by revenue.

Amazon and Flipkart vs your own website: a channel profit guide for Indian D2C brands, cover

Indian D2C founders argue about this constantly, and usually with the wrong numbers. One side points at marketplace revenue. The other points at website ROAS. Neither figure tells you which channel makes money. A marketplace order and a website order with the same selling price can leave very different amounts in the bank once fees, shipping, returns, advertising and payment costs are taken out. This guide shows how to compare the two properly, what each channel is good for, and when to shift the mix. It builds on our guide to [contribution margin for D2C brands](/insights/d2c-contribution-margin-cm1-cm2-cm3), which explains the margin layers used here.

What each channel really gives you

Marketplaces and your own website do different jobs. Treat them as complements before you treat them as rivals.

Marketplaces give you

  • Existing demand. People open Amazon or Flipkart intending to buy. You are meeting demand, not creating it.
  • Trust and delivery. Buyers trust the platform's payment, delivery and returns, which helps a brand nobody has heard of.
  • Reach beyond metros. Marketplace logistics reach pin codes that a young brand's own courier arrangements may serve poorly.

Marketplaces take away

  • The customer relationship. You get limited customer data and few ways to contact buyers afterwards.
  • Pricing freedom. Your listing sits beside competitors and is compared on price and rating.
  • Control. Fees, search ranking rules and policies can change, and your listing can be suspended.

Your own website gives you

  • Customer data and consent. Email, phone and purchase history you can use for retention, within the law. See our note on the DPDP Act for marketing teams.
  • Control of brand and offer. Bundles, subscriptions, content and pricing are yours to design.
  • Better signals for advertising. With proper server-side tracking, your ad platforms learn from real orders.

Your own website costs you

  • Traffic. Every visitor has to be earned or bought.
  • Operations. Checkout, payments, shipping, returns and customer service are all your responsibility.
ROAS looks great. The bank balance does not. Why D2C revenue in a dashboard is not money made, and how contribution margin by product changes where the budget should go. 51 seconds. Narrated with an AI voice. Charts shown are illustrative. See fractional CMO for D2C brands
Read the transcript

Return on ad spend looks great. The bank balance does not. For D2C brands, revenue in a dashboard is not the same as money made.

Return on ad spend counts orders. It ignores what they cost you. COD returns and discounts rarely reach the dashboard. Growth can look healthy and still drain cash.

Stop optimising for revenue. Start with margin.

See CM1, CM2 and CM3 by product and channel. Move budget towards what actually makes money. Nudge COD to prepaid and report delivered orders.

D2C performance marketing, with Rahul D Sarker. Book a call at rahuldsarker.co.

Every cost on each side

Most comparisons go wrong because one channel's costs are counted in full and the other's are not. List them side by side. Fee structures change, so take current figures from the official Amazon fee schedule and Flipkart fee page for your category, and from your own payment, courier and agency contracts.

Costs to include for each channel
CostMarketplaceOwn website
Platform feesReferral or commission fee, closing or fixed fee, and other category chargesStore platform subscription and app costs
FulfilmentMarketplace fulfilment and storage fees, or your own shipping to their standardsCourier charges, packaging and warehouse costs
PaymentsIncluded in platform fees in most casesPayment gateway charges; cash on delivery handling fees
ReturnsReturn shipping, refund handling, damaged or unsaleable stockReturn to origin on refused COD orders, reverse logistics, refunds
AdvertisingSponsored product and brand ads inside the marketplaceMeta, Google and other paid traffic; creative production
DiscountsPlatform sale events, coupons and deal participationYour own offers, first-order discounts and free shipping thresholds
People and toolsListing management, catalogue, marketplace agency feesSite development, CRM and messaging tools, customer service
Tax handlingTax collected or deducted by the platform, reconciliation effortGST on your own invoices

Ask your accountant how platform tax deductions and credits affect cash flow. They do not change profit, but they change when you receive money.

How to calculate profit per order by channel

Work with real orders from the last three months, not with assumptions.

  1. Start with net revenue per order. Selling price less discounts and less GST.
  2. Subtract product cost. Landed cost of goods, including packaging.
  3. Subtract channel fees. Marketplace fees, or payment gateway and platform costs for the website.
  4. Subtract fulfilment. Forward shipping and storage.
  5. Subtract the cost of returns, spread across all orders. Total return and RTO cost for the period divided by the number of orders dispatched. This is the figure most often missed. Our article on RTO-adjusted ROAS explains why delivered orders, not placed orders, are the right base, and the RTO-adjusted ROAS calculator does the arithmetic.
  6. Subtract advertising per order. Channel ad spend divided by delivered orders from that channel.
  7. What remains is contribution per order. Compare the two channels on this figure and on contribution as a share of net revenue.

Then add one more view: profit per customer over six or twelve months. A website customer you can contact may place several more orders at low cost. A marketplace customer may reorder too, but you pay fees and often advertising on every order. Channel decisions made on first-order profit alone tend to undervalue the website.

Advertising works differently on each side

Marketplace advertising captures demand that already exists. Someone searches for a product type and your sponsored listing appears. It is close to the purchase, so reported returns often look strong. Two cautions apply. Some of those sales would have happened anyway through your organic listing, particularly on searches for your own brand name. And advertising, reviews and sales rank reinforce each other, so pausing ads can affect organic position.

Website advertising mostly creates demand. Meta and Google campaigns introduce the brand to people who were not searching for it. This costs more per order but builds an audience that marketplaces also benefit from: many shoppers see an advertisement, then search for the brand on a marketplace to buy with familiar delivery and payment. That spillover means website ad spend is partly funding marketplace sales, and neither channel's own report shows it.

For this reason, judge the business on blended figures as well as channel figures. MER, which compares total revenue with total marketing spend, shows whether the whole system is working. Where budgets allow, a simple holdout test will tell you how much of your marketplace revenue depends on off-platform advertising.

The retention difference

The largest long-term difference between the channels is what happens after the first order.

On your own site, you can build a second-order programme: post-purchase messages, replenishment reminders, bundles, loyalty benefits and win-back campaigns through email and WhatsApp. For one supplement brand we worked with, retention revenue rose from 22% to 41% of sales over an eight-month engagement that rebuilt email flows around purchase history and added a loyalty programme (case study).

On a marketplace, your retention tools are limited to the product itself, packaging, reviews and whatever follow-up the platform allows. Use package inserts carefully and within each marketplace's policies, which restrict how sellers may contact or redirect buyers.

If your product is bought once, such as a durable item, this difference matters less. If it is consumed and reordered, it is the main argument for building the website channel even when first-order profit is lower.

When to lean on each channel

Use this as a starting point and test it against your own numbers.

Which channel to lead with
SituationLead withWhy
New brand with little awarenessMarketplace, with a simple websiteBorrowed trust and existing demand while you learn what sells
Consumable product with repeat purchaseWebsite, supported by marketplaceRepeat orders are worth far more when you own the relationship
Low-priced item with high shipping shareMarketplacePlatform logistics and basket-building make small orders viable
High COD share and high returns on the siteFix website returns firstShifting budget to the site before fixing RTO loses money faster
Premium or story-led productWebsiteYou need space for content and control of presentation
Heavy price competition from lookalike sellersWebsite, plus brand registration on marketplacesDifferentiation is hard on a comparison shelf
Urgent, small-basket purchases in citiesQuick commerce alongside bothA different buying occasion altogether

Quick commerce has its own economics and advertising. See our guide to ads on Blinkit, Zepto and Instamart.

Running both without channel conflict

Most established brands sell through both. The work is in stopping them undermining each other.

  • Keep pricing consistent. Large price gaps between channels train customers to wait and can breach marketplace policies. Differentiate with bundles, sizes and exclusives instead.
  • Give the website something the marketplace does not have. Subscriptions, larger packs, combinations, early access and loyalty benefits.
  • Use marketplaces for discovery, the website for depth. A hero product range on marketplaces and the full range on your site is a common pattern.
  • Hold one view of stock. Overselling during sale events damages ratings on both sides.
  • Protect your brand search. Bid on your own brand name where competitors do, on both search engines and marketplaces.
  • Report on one sheet. Revenue, contribution and marketing cost by channel, with a blended total. Our weekly performance report template can be extended with a channel view.

Signals that it is time to shift the mix

Review the mix every quarter. Move budget towards the website when:

  • Website contribution per order is positive after returns and improving.
  • Repeat purchase rate on the site is rising.
  • Marketplace fees or advertising costs are taking a growing share of revenue.
  • Tracking and attribution are sound enough to trust the numbers. A marketing attribution review is worth doing before a large shift.

Move budget towards marketplaces when:

  • Website acquisition costs are rising faster than order values.
  • COD refusals and returns on the site remain high despite fixes. Our guide to moving COD buyers to prepaid covers the options.
  • A category is growing quickly on a platform and you are under-represented there.

Make changes in steps, and watch blended revenue and contribution for a few weeks after each one, because the channels influence each other.

A 30-day channel audit

  1. Week 1: gather. Export three months of orders, fees, shipping, returns and ad spend for each channel.
  2. Week 2: calculate. Work out contribution per order and per customer for each channel using the method above.
  3. Week 3: diagnose. Find the two or three costs that most reduce contribution on each side. Check tracking and the share of marketplace sales on brand searches.
  4. Week 4: decide. Set the budget split for the next quarter, the fixes to make and the measures to watch.

If the numbers are hard to assemble, that is itself the finding: the business is being run without a view of profit by channel. Our D2C growth marketing work usually starts there, and a free 30-minute consultation is enough to see where your largest leak is.

FAQ

Most new brands benefit from starting on marketplaces for demand and trust, with a simple website alongside to begin collecting customers and data. Shift emphasis as you learn which products sell and whether customers reorder.

It depends on your category, order value, return rates and advertising costs. Calculate contribution per order and per customer for each channel from your own data. Revenue or ROAS alone will mislead you.

Only within each marketplace's policies, which limit how sellers contact or redirect buyers. The safer approach is to build brand recognition through the product, packaging and advertising so customers look for you directly.

Some sponsored sales would have happened anyway, especially on searches for your brand name. Test by adjusting brand-term bids and watching total sales, not only advertised sales.

Start from contribution per order and repeat purchase behaviour on each channel, then test changes in steps. Judge the result on blended revenue and contribution, since website advertising often lifts marketplace sales too.

Read this article on your favourite platform

Ready to build the system?

If this describes your funnel, a 30-minute call will find where the constraint sits in your own numbers and what it would take to fix it.

It starts with a 30-minute call. Pick a time below.

Prefer email?