Amazon’s New Cancellation Fee Has a Tier You Didn’t Check
Stockout on Amazon. SKU already sold out on Shopify. Order auto-cancels after 24 hours.
₹0 charged — only a Cancellation Rate metric hit.
What actually changed on 17 August
Amazon India didn’t raise its cancellation fee. It rebuilt it. The old fee was pegged to the referral-fee rate of whatever category the product sat in — apparel, electronics, home goods each had a different, indirect number that few sellers could recite from memory. Effective 17 August 2026, Amazon replaced that with a tiered fee based on order value, applied to every Easy Ship and Self Ship seller.
The headline most sellers read was “up to 10%.” That’s true, but it’s the top of a ladder, not a flat rate — and where you land on that ladder decides whether this is a rounding error or a real hit to your margin.
| Order value band | Cancellation fee | + GST (18%) | Effective rate |
|---|---|---|---|
| Up to ₹10,000 | 10% | on the fee amount | 11.8% of order value |
| ₹10,001 – ₹50,000 | 8% | on the fee amount | 9.44% of order value |
| ₹50,001 – ₹1,00,000 | 5% | on the fee amount | 5.9% of order value |
| Above ₹1,00,000 | 2% | on the fee amount | 2.36% of order value |
Source: Inc42 fee structure report, corroborated by Amazon’s own seller forum notice.
If you sell apparel, beauty, home decor or accessories under ₹10,000 a ticket — which is most of the Indian D2C catalogue — you sit in the top tier by default. That’s not an edge case. That’s your whole SKU list.
The 18% GST is charged on the fee, not folded into the percentage. A 10% fee on a ₹1,000 order is ₹100 — plus ₹18 GST on that ₹100 — for a ₹118 total, an effective 11.8% of order value. Some secondary write-ups have floated higher, uncorroborated figures for this stack; the tier table above is the one confirmed against Amazon’s own forum notice, and it’s the one to plan around.
The part everyone skips: it fires automatically
Here’s the sentence in the policy that changes the calculus for a lean-ops seller: the fee applies not just when you choose to cancel an order, but when the order auto-cancels because you didn’t confirm shipment within 24 hours of the estimated ship date. Miss that window — because a SKU sold out on another channel, because your OMS sync ran late, because nobody checked the dashboard over a weekend — and the system treats it exactly like a deliberate cancellation. Same fee, same tier, no manual override required.
Flip the switch above: the exact same stockout costs ₹0 before 17 August and ₹106.20 after it — same mistake, same order, different bill.
Amazon’s own position, given to press, is that this affects “less than 1% of orders” on the platform — a company framing worth noting, not a neutral third-party audit. Small in aggregate doesn’t mean small for the seller it happens to. A <1% platform-wide rate can still be a routine, repeat cost for any seller whose inventory sync isn’t tight, because cancellations cluster where the operational weakness is — not evenly across the marketplace.
Amazon frames this as affecting under 1% of orders platform-wide. That’s an aggregate across every seller on Amazon.in — not a claim about any single seller’s exposure. If your SKUs run thin inventory across multiple channels, your personal rate can sit well above the platform average, because cancellations concentrate exactly where sync is weakest.
What it costs you, in rupees
Two worked examples, both drawn from the tier table above.
₹899 kurta, auto-cancelled — 10% tier
- ₹899 × 10% = ₹90 fee
- ₹90 × 18% GST = ₹16.20
₹106.20 per incident
That looks trivial as a single line item. It isn’t at scale — 40 such incidents a month through festive ramp-up is roughly ₹4,250/month in fee leakage from a failure mode nobody deliberately chose.
₹45,000 furniture order, seller-cancelled after two missed pickups — 8% tier
- ₹45,000 × 8% = ₹3,600 fee
- ₹3,600 × 18% GST = ₹648
₹4,248 on a single order
The policy as published has no stated carve-out for third-party courier failure. If the pickup partner doesn’t show up twice and you cancel to avoid breaching your SLA window, the fee applies the same as if you’d changed your mind.
Formula: (order value × tier %) × 1.18 × cancellations per month
It’s not just Amazon, and it’s not staying still
This fee didn’t land in isolation. It’s the first of three moves in a four-week window, all landing before India’s biggest selling season.
10/8/5/2% + GST for Easy Ship and Self Ship sellers.
₹30 missed Dispatch-By-Date, ₹60 seller-cancelled or repeat miss, ₹90 delayed-then-cancelled. Applies to sellers active 3+ months.
Forum commenters push back citing cancellations “beyond their control.”
+₹1 on items ≤₹500, +₹3 above — a second, compounding cost change inside three weeks.
The single heaviest order-volume period of the year — every fee above compounds fastest here.
Run the same item on both marketplaces and miss a dispatch window on both in the same week, and you don’t pay one fee — you pay Amazon’s tiered cancellation fee and Flipkart’s flat per-shipment penalty for what was a single underlying failure, like a warehouse short-staffed for a day. The two policies don’t substitute for each other. They stack.
The framework: where does your exposure actually come from
Before you can reduce this cost, you need to know which of three failure types is driving it — because the fix for each is different, and only one of them is something a listing or photography fix can touch.
- Inventory-sync failure — SKU sells out on another channel before Amazon’s stock count updates. Fix: tighter OMS sync frequency, buffer stock on fast-movers.
- Logistics/courier failure — pickup partner misses the window, forcing a seller-side cancellation to avoid an SLA breach. Fix: courier SLA renegotiation, backup pickup partner.
- Variant/pack confusion — buyer orders the wrong size/pack because the primary image doesn’t show it clearly, triggering a return-as-cancellation flow. Fix: visual and copy clarity on the hero image and gallery, not just inventory ops.
Is the cancellation showing up mostly on SKUs with multi-channel inventory (Amazon + Shopify + Flipkart)?
This matters because the temptation, reading a fee like this, is to throw the whole budget at inventory software. For a lot of sellers, a meaningful slice of the cancellation-triggering confusion is sitting in the product photography itself — a pack size that’s legible in the title but invisible in the hero image, so the wrong variant ships and comes back as a cancellation-coded return.
The pre-festive checklist
Seven concrete, ownable actions to take before October — not just “be more careful.”
0 of 7 done
The fee itself isn’t the story — ₹106 on a stockout, ₹4,248 on a courier failure, these are line items, not existential threats. The story is that a marketplace just converted an operational weakness you could previously ignore into a metered cost that scales with your order value and your mistake rate, right before the quarter where both go up. You don’t need to renegotiate with Amazon to change your number here. You need to know which of the three causes — sync, logistics, or listing clarity — is actually generating your cancellations, because two of those are ops problems and one of them is sitting in your product photography.
Next step: pull your last 60 days of cancelled orders this week, tag each by cause using the framework above, and if variant or pack-size confusion shows up as a repeat pattern, start with a visual conversion checklist audit on those specific SKUs before October volume arrives.
- New Amazon.in cancellation fee tiers (10/8/5/2% + 18% GST), effective 17 Aug 2026 — Inc42
- Confirmation of tiers, GST treatment, auto-cancellation trigger, and old-vs-new fee basis — Amazon Seller Central Forum
- Amazon’s “<1% of orders” company statement — Free Press Journal
- Amazon closing-fee increase (effective 7 Sep 2026) and Flipkart’s ₹30/₹60/₹90 penalty structure (effective 23 Aug 2026) — Inc42
No. The policy as published is scoped to Easy Ship and Self Ship sellers — the two models where Amazon doesn’t control fulfilment directly. FBA sellers are not the stated target of this specific change since Amazon handles storage and shipping for those orders. If you run a mixed FBA/Easy Ship catalogue, only the Easy Ship/Self Ship portion is exposed to this fee.
No, and this is the most common misreading. It’s a four-tier structure: 10% under ₹10,000, 8% for ₹10,001–₹50,000, 5% for ₹50,001–₹1,00,000, and 2% above that — plus 18% GST charged on top of the fee amount, not folded into it. Most apparel, beauty and accessories sellers land in the top tier because their average order value sits under ₹10,000.
No. The policy explicitly excludes buyer-initiated cancellations. It applies to seller-initiated cancellations and to auto-cancellations that happen when a seller fails to confirm shipment within 24 hours of the estimated ship date — including ordinary stockouts, even unintentional ones.
No documented carve-out for third-party logistics failure appears in the policy as reported. A seller who cancels because a courier missed pickup twice is charged the same as one who changed their mind. If you rely heavily on a single courier partner, a backup arrangement is worth pricing against the fee exposure now, not after a bad week in November.
It doesn’t offset it — the two stack. Flipkart’s separate fulfilment-failure penalty (₹30–₹90 per shipment, effective 23 August 2026) applies to sellers active three or more months and is charged independently of Amazon’s fee. A single operational failure — a missed pickup on an item listed on both platforms — can trigger both bills in the same week.
Audit your primary product images for variant and pack-size clarity. A share of seller-initiated cancellations trace back to wrong-item or wrong-variant confusion that starts with an unclear hero image, not a broken inventory system — and that’s a photography and listing fix, not an ops overhaul, so it’s the fastest lever to pull before festive volume arrives.
Get your listings festive-ready before the fee finds them
A share of seller-initiated cancellations trace back to unclear variant and pack-size photography, not broken inventory systems. Start with a specific read on your own listings before October volume triples.
