Channel conflict is not solved by policing discounts, it is solved by giving each channel a different job and a different pack to sell. Work out contribution per unit on every channel after commission, fulfilment, returns and ad spend, hold a single MRP for any given pack across Amazon, Nykaa, Flipkart and your own site, and differentiate on bundle, size and format rather than on price. Your own website should match the marketplace number and win on range, early access and service, because that is the one advantage a platform cannot copy.
- →Channel conflict is caused by having no architecture, not by having the wrong discount. Decide what each channel is for before you touch a price.
- →Rank channels on contribution per unit after commission, fulfilment, returns and ad spend. Gross margin on MRP flatters every channel equally.
- →Hold one MRP per pack everywhere and differentiate on pack size, bundle and format instead. Two channels selling different things cannot be price compared.
- →Your own site should match the marketplace price and win on range, early access, refills and service. Undercutting trains your best customers to wait.
- →Budget sale event participation in advance. Deep in one planned event beats shallow discounts across every event, which permanently resets your price.
Channel Conflict Is an Architecture Problem, Not a Discount Problem
Most founders meet channel conflict as a fire. Someone screenshots the same jar selling for less on a marketplace than on the brand's own site, a distributor calls, and the reflex is to fix the number on whichever channel looks wrong this week. That is firefighting, and it never holds, because the number was only ever the symptom. The underlying problem is that the brand never decided what each channel is for, so every channel is competing to sell the identical thing to the identical buyer, and price is the only lever any of them has left.
An architecture is the alternative. It says, before anything goes live, which SKUs exist on which channel, what a buyer gets at each price point, who is allowed to sell, and what happens during a sale event. Once that exists, most conflict disappears without anyone changing a price, because the channels stop being interchangeable. The brands that never get to this point end up managing a permanent race to the bottom that they themselves started.
The Indian context makes this sharper than it is elsewhere. A printed MRP sets a ceiling every channel works below, marketplaces fund their own discounts out of their margin whether you like it or not, and shoppers are unusually comfortable checking three apps before they buy. You cannot control the final displayed price on a platform you do not own. What you can control is the structure underneath it, which is the price you sell in at, how many hands the product passes through, and whether the same exact pack is even available in both places.
It also helps to be honest about why you are on each channel. Marketplaces buy you discovery and default trust. Your own site buys you margin, data and the ability to talk to a customer again. Quick commerce buys you convenience purchases you would otherwise lose. Those are different jobs, and a pricing architecture is simply the decision to stop pretending they are the same one.
Work Out What Each Channel Actually Pays You
Before any pricing decision, you need contribution per unit per channel, and most brands do not have it. They have gross margin on MRP, which is a number that flatters every channel equally and tells you nothing about which one deserves inventory. The honest calculation starts from what the customer actually pays after coupons and bank offers, then removes everything the channel takes on the way back to you.
Do it once per channel and the ranking is often not what the team assumed. A marketplace with a high commission can still out-earn your own site if the site's paid acquisition cost is carrying the whole order. A channel that looks efficient on paper can turn negative once returns are counted properly, which is a familiar story in fashion, where the return rate quietly decides whether a category is profitable at all. Averages hide this, so run the numbers at SKU level, not category level.
The line item brands most often leave out is visibility. On a mature marketplace you are not choosing between advertising and not advertising, you are choosing between paying for placement and being invisible, so ad spend belongs inside channel cost rather than in a separate marketing budget. Once it sits there, decisions about how much of the budget a sponsored campaign should really be absorbing become pricing decisions rather than media decisions, which is what they always were.
- Start from the shelf price, not the MRP What the customer actually pays on that channel once coupons, bank offers and platform discounts are applied.
- Deduct the channel take Commission, closing fee and payment charges, which differ by platform and by category.
- Deduct fulfilment and returns Weight-slabbed shipping, storage, and the cost of returns and RTO travelling in both directions.
- Deduct the visibility spend Sponsored ads, event participation and merchandising, which are not optional if you want the traffic.
- Compare what is left, channel by channel Rank channels on contribution per unit, then set discount policy on that ranking rather than on shelf price.
Hold One MRP and Differentiate on What You Bundle
The rule that solves most channel conflict is simple to state and uncomfortable to follow: one MRP for one pack, everywhere. The moment the identical pack carries a different price on two channels, you have taught the customer to treat your brand as a commodity and to shop for the cheapest window. Holding one MRP does not mean every channel sells at the same final number, because platform funded promotions will move it. It means the difference is never something you authored.
What varies instead is the offer. Give each channel a pack it can own. A larger value pack for a marketplace where basket size and free delivery thresholds drive the purchase. A trial or travel size where discovery buyers dominate. A refill format on your own site where the customer already knows the product. A festive pairing that exists for six weeks and then does not. Two channels selling different things cannot be compared on price, which is the entire point, and it is a far cheaper solution than subsidising a permanent discount.
Category behaviour should decide which pack goes where. Beauty buyers arriving on a specialist platform browse by concern and are comfortable with sets, while the same buyer on a general marketplace is often replenishing something they already use, which is a real difference in intent rather than a difference in price sensitivity, and it is the same reasoning behind choosing where a skincare range should lead and where it should merely be present. Match the pack to the intent and the price question quietly stops being the interesting one.
This also needs a floor, written down and shared with everyone who sells you on. Decide the lowest selling price you will support on each channel, set your sell-in price so no partner can go under it without eating their own margin, and keep the seller count on your listings small enough that you know who they are. A floor that lives only in a founder's head is not a floor.
What Your Own Site Should Win On, and It Is Not Price
Founders often want their own website to be the cheapest place to buy, on the logic that it is the channel with the best margin. It backfires twice. It trains your most loyal customers, the people who came looking for you by name, to wait for the cheaper number, and it gives marketplace pricing systems a reason to treat your listing as uncompetitive. You end up paying for the privilege of undercutting yourself.
Price parity plus a better deal is the stronger position. Same number, more around it. Full range including the sizes and shades no marketplace will stock. New launches first, for a few weeks, before they go wide. Subscription or refill pricing that rewards repeat purchase rather than first purchase. Samples chosen for what the customer already bought. Direct service when something goes wrong. None of these are discounts, and none of them can be copied by a channel that does not know who the customer is.
The site's real advantage is that the relationship continues after the order, which is exactly what a marketplace withholds. A buyer you can reach again is worth more than a buyer you cannot, and the practical version of that is a working retention loop rather than a louder acquisition budget, which is why owning the repeat purchase conversation directly tends to do more for site economics than any price cut. Judge the site on repeat rate and contribution per customer, not on whether it beat Amazon on a single order.
| Marketplace listing (Amazon, Nykaa, Flipkart) | Your own site, at the same price |
|---|---|
| Discovery and default trust from a platform the shopper already opens | The full range, including sizes and shades no marketplace will stock |
| Delivery speed and a saved payment method you cannot match | New launches first, for a few weeks before they go wide |
| The hero SKU and the pack sizes that sell fastest | Refill formats and subscription pricing that reward repeat purchase |
| Promotions the platform funds and schedules, on its calendar | Samples chosen for what this customer already bought |
| The platform keeps the customer, the contact and the purchase history | A customer you can reach again, and serve directly when it goes wrong |
Surviving Sale Events Without Resetting Your Floor
Sale events are where carefully built architectures get dismantled in a fortnight. The platform proposes a discount band, the category manager implies visibility depends on accepting it, and the brand agrees because the alternative is disappearing during the highest traffic week of the year. Then the event ends, the price goes back up, sales fall off a cliff, and everyone concludes the brand only sells on discount. It usually does, by then.
Decide before the calendar arrives which events you will participate in and at what depth, and treat that as a budget with a number attached rather than a series of individual negotiations. Participation is a real cost, so it should compete with your other spend on the same terms. Going deep in one event you have stocked and merchandised for beats appearing at a shallow discount in every event of the year, which is the pattern that slowly resets a customer's idea of what your product is worth.
Protect the floor with the offer, not the price, wherever you can. Bundles, larger packs and gift-with-purchase move the perceived value without moving the number on your core SKU, and they leave you somewhere to return to when the event ends. Where a platform funds the discount out of its own margin, take it if the volume is worth the expectation you are creating, and pass when it is not. Recovery after the event matters too: plan the two weeks afterwards deliberately, because a listing that goes quiet immediately after a sale loses the ranking the sale bought.
"You cannot control what a marketplace displays. You can control what it is selling, who is allowed to sell it, and what you refuse to sell it for."
- Brand Integer Marketplace Team
Frequently Asked Questions
Should we keep the same price on Amazon, Nykaa and our own website?
Keep the same MRP everywhere and let the selling price differ only where the channel's own economics or a platform funded promotion force it to. Customers cross check, and a product that is visibly cheaper on one marketplace teaches everyone to buy there permanently, which hands your margin and your customer data to that platform. What should differ between channels is not the number on the same SKU but what a buyer gets for it: pack size, bundle composition, a variant that only exists in one place. That way each channel has a distinct offer to make and nothing looks like a discount on the identical thing.
Can we stop a marketplace from discounting our product below our floor?
Not directly, because the seller of record sets the selling price and platforms fund their own promotions. What you can control is upstream. Set the distributor and seller price so that nobody underneath you has room to go below your floor without losing money themselves, keep the number of sellers on your listing small and known, and use brand registry controls so unauthorised sellers can be challenged. Where a platform funds a discount out of its own margin, decide whether the volume is worth the reset in customer expectation before you agree to be included in the event.
Should our own website undercut the marketplaces to pull customers across?
It is tempting because the site looks like the cheapest channel to serve, and it is usually the wrong move. Undercutting trains your best customers to expect a lower number and gives marketplace algorithms a reason to demote you for price inconsistency, while doing nothing about the reason people default to Amazon, which is delivery speed and a saved payment method. Match the price and win on things a marketplace structurally cannot copy: full range availability, first access to new launches, a subscription, refills, samples and direct service.
Does listing on several marketplaces at once cannibalise sales?
Rarely in the way founders fear. Shoppers in India are largely platform loyal, so a customer who buys beauty on Nykaa is not usually the same person who would have bought it on Amazon. What actually cannibalises is inconsistent pricing across those platforms, because it converts a range decision into a price comparison. The real cost of adding a channel is operational, not demand side: more inventory to hold, more content to maintain and more returns to process, which is why each channel should earn its place on contribution rather than on the fact that it exists.
