Leaving a marketplace: the order of operations
Switching off your listings is the first thing most brands do when they leave a platform, and it is usually the most expensive thing they could have done. The stock is still in the platform's building, the returns window has weeks to run, and you have just removed the page that would have sold the stock.
- Delisting first strands your stock inside the platform's network with no sales page left to clear it through, so run the inventory down before you touch visibility.
- The returns window outlives your last order, which means customer service, refunds and reverse logistics have to stay staffed for weeks after you stop selling.
- Ratings, review count and listing history do not travel with you and cannot be reclaimed if you return later, which is the true cost of a full exit.
- Going dormant is not free because storage, account fees and a decaying seller rating continue, so decide deliberately between dormant and gone.
Exiting a platform is a sequencing problem more than a negotiation. The order you do things in decides whether you leave with your cash or leave with a warehouse full of returns and a settlement nobody is watching. Note that this is a departure, not a transfer: if you are moving the same business to a new legal entity, the mechanics are different and changing your selling entity covers that case.
Turning off the listings first is the expensive move
The instinct is to kill visibility on day one so the losses stop. It does the opposite. Your sales page is the cheapest route to clear the stock that is already sitting inside the platform’s network, and once the listing is dark you are left with recall fees, freight and a pile of units to place somewhere else. Sell down first, then reduce visibility, then delist.
A workable order looks like this. Stop inbound shipments and cancel open purchase orders into the platform. Stop the ad spend that was only defending that channel. Let organic demand and a controlled markdown run the stock down. Recall what is left. Handle open orders and the returns tail. Close the listings last. Any markdown you use here should be planned like any other run-down, not improvised, and markdown timing and recovery rate is the right lens for it.
Your stock is in their building
Inventory in a fulfilment centre or a dark store is the constraint that sets your exit date, not your intention. Removal is charged per unit, processed on the platform’s timetable rather than yours, and comes back in worse condition than it went in. Expect shortfalls, mixed cartons and units that fail QC on receipt. Long term storage fees keep accruing while you wait, which is exactly the trap described in long term storage fees, so a slow exit is not a free exit.
Do the arithmetic before you decide to recall at all. For low value units, the removal fee plus freight plus grading plus the cost of finding a new home frequently exceeds what the stock is worth. Selling it out on the platform at a discount, or disposing of it there, often nets more. What comes back needs a plan too, and liquidating dead stock is a better route than letting it age in your own warehouse instead of theirs.
The tail runs longer than your last order
Your last sale is not your last obligation. The returns window on that order runs for its full duration, and on quick commerce and marketplace platforms the disputes, damage claims and quality complaints attached to it run longer. That means reverse logistics capacity, a grading process and someone to make disposition calls stay live after the channel is closed. Keep the returns disposition process running rather than quietly disbanding it.
Money behaves the same way. Reserves release on a schedule, deductions land after your final invoice, and reimbursement claims you filed in the last month resolve weeks later. Keep the account, the bank mandate and a named owner for reconciliation alive until the ledger closes, using the same discipline as your normal settlement reconciliation. Brands that hand the login back too early simply forfeit the balance.
The customer still thinks you sell there
Someone who bought from you on that platform will come back to it looking for you, and will find nothing. They will then contact you directly, or complain publicly, or assume the brand has disappeared. Decide in advance where those people land: a support address that stays monitored, a page on your own site that says where to buy now, and a briefed customer service team that can handle a warranty or refill request for an order placed on a channel you no longer operate.
Reviews and history are the real switching cost
Fees and freight are recoverable. Ratings are not. Review count, star average, listing age, search rank built on months of order velocity and your seller rating and buy box history stay with the platform. If you come back in two years you come back as a new seller against competitors who never left, and rebuilding that position costs far more than the fees you saved. Reviews are an asset you have been accumulating, which is the argument in treating reviews as an operations function, and a full exit writes that asset off.
Dormant, or actually gone
Dormant means listings stay live with zero stock or hidden, the account stays open, and the history survives. It buys you an option on returning. It also costs: account fees where they apply, storage on anything left behind, a rating that decays without orders, and a listing nobody is watching, which is how suppression and hijacking happen. Dormant is a real choice, but only with a named owner and a review date. Dormant by neglect is the worst of both.
Tell the category manager yourself
A category manager who learns you have gone from a dashboard is a relationship you have ended without meaning to, and Indian category teams move between platforms constantly. Tell them before the stock recall goes in, give a straight reason, and say whether the door is open. In practice this also buys you cooperation on the things you still need from them: removal timelines, deduction disputes and a clean settlement. The same courtesy applies downstream to distributors and any offline partner affected, which is the point of telling the channel before a change lands. If the decision itself is not settled yet, work through the channel exit trigger first.