On June 29, Amazon will start setting the handling time itself for any seller-fulfilled SKU that keeps shipping faster than it promises. The seller loses control of that listing's delivery date.
How it works: Handling time is how long a seller tells Amazon it takes to get an order to the carrier. Amazon uses it to build the delivery date a shopper sees, and that date helps decide who wins the Buy Box.
A SKU gets flagged "inaccurate" when it ships at least a day faster than its set handling time, consistently, over 30 days, and the seller then gets 30 days to fix it.
After that, Amazon takes over the setting, bases it on the seller's recent ship speed, and adds 180 days of protection from late-shipment penalties. Custom, handmade, and Heavy & Bulky LTL items are exempt.
Amazon's reason is simple: it says a one-day-faster promise lifts sales about 5%.
Why it matters: Sellers pad handling time on purpose. The extra day or two is a buffer for busy weeks, a missed pickup, or a weekend with no one shipping.
Amazon now reads that padding as inaccurate and moves the promise up to match how fast the seller really ships, so once the 180-day grace period ends, the seller has to hit the tighter date every day.
The mechanic is the part to watch. Automated Handling Time uses a rolling average, so a run of fast shipping pulls the bar down, and a normal day after it can count as late. Steady shipping is safer than fast one week and slow the next.
The economics: A faster promise costs money to hold: weekend labor, earlier carrier cutoffs, stock kept closer to customers.
Those costs land whether or not the extra speed sells more. On high-margin, fast-moving SKUs, the 5% lift usually covers it, and the tighter promise is worth keeping. On low-margin slow movers, it often does not. Channel is the other lever. Stay seller-fulfilled and you keep both the margin and the risk; move the SKU to FBA and Amazon owns the speed, for a fee.
The rule quietly makes seller-fulfilled more expensive on anything a seller cannot ship fast and consistently.
A few things to weigh: Not every SKU is at risk. The ones to watch ship faster than they promise, month after month, and those are the listings Amazon will reset. Where that gap exists, the promise tightens either way, and the only question is whether the seller picks the number or lets Amazon's average pick it.
If a SKU genuinely cannot go faster, the common levers are a weekend ship shift, inventory held regionally, an earlier order cutoff, or a move to FBA. One reaction going around the seller forums, holding orders back so the average does not drop, is a trap, since it is slower for the buyer and still risks penalties on a busy day.
For a seller who already ships fast and steady, this barely registers; it bites hardest where the promise carried more cushion than the operation needed.
Across the sector: This is Amazon's second accuracy crackdown in 2026, after February's move to drop weak listings one at a time rather than across a whole account. Walmart is doing the same kind of thing, with a new eight-metric seller scorecard.
On both, the delivery promise is now graded listing by listing, and a seller's visibility rides on keeping it accurate. Setting it once and forgetting it no longer works.
Dig deeper:
How the Amazon Buy Box works - BigCommerce [Explainer] - why ~82% of sales flow through it.
How Amazon's On-Time Delivery Rate is calculated - Riverbend Consulting [Explainer] - the 90% on-time threshold behind the change.
