Sea-Intelligence released issue 180 of its Global Liner Performance report on August 26. Global schedule reliability for July came in at 56.4%. That is down 6.1 percentage points from June and 8.8 points from July 2025, and it is the weakest month of 2026. (Container News, August 30)
The second number matters more. Vessels that arrived late arrived an average of 6.06 days late. That is up 0.59 days from June and 1.34 days from a year ago.
Read them together. Just over half of sailings hit their published window. The ones that miss now miss by most of a week.
Rates went down the same week reliability did
Drewry published its World Container Index on August 27. The composite fell 1% to $4,473 per 40ft container. Shanghai to New York dropped 2% to $9,333. Shanghai to Los Angeles held flat at $6,818. (Drewry World Container Index)
So a landed-cost spreadsheet looks fine right now. The price per container is behaving. The number that moved is the arrival date, and most brands do not track that one at all.
What is causing it
Linerlytica counted 4.31 million TEU of capacity stranded by port congestion, reported August 25. That passes the previous record of 4 million TEU set during the 2022 post-pandemic surge. More than 1.5 million TEU is sitting at anchor off Shanghai alone. Typhoons through Asia pushed Chinese port operations off schedule, and the backlog spread to Singapore, Busan, Colombo, and Mumbai. (The Maritime Executive, August 25)
One caveat belongs with that headline. The container fleet is far larger than it was four years ago, so 4.31 million TEU represents 12.6% of global capacity today against 15.7% in 2022. The absolute figure is a record. The share of the fleet is not. That second number is why this reads as a timing problem rather than a capacity shortage.
The part that hits your warehouse
Late is the obvious problem. Bunched is the expensive one.
When reliability runs at 90%, containers land roughly when you told your 3PL they would land, and receiving spreads across the month. At 56.4%, arrivals cluster. Four purchase orders you deliberately spaced across a month can show up inside the same week because two of them slipped and the others did not.
That is a receiving problem more than a freight problem. Your dock has a fixed throughput. Detention and demurrage clocks start whether or not there is labor available to unload. Every day a pallet sits unreceived is a day that inventory is not sellable, and in October that is the only measure that counts.
What this means for your brand
If you import and you ship 1,000 to 25,000 orders a month, the assumption to change is your buffer, not your carrier.
Two numbers are worth pulling this week.
Your actual arrival variance. Take your last six inbound containers. Compare the ETA you were quoted at booking against the day the container was genuinely available to receive. Do not use the carrier's revised ETA, because that number gets rewritten every time the schedule slips. Use the original. If your average slip lands anywhere near six days, your holiday inbound cutoffs are already wrong.
Your receiving capacity in units per day. Most brands have never asked their 3PL for this figure. You need it because the live risk this quarter is two containers arriving inside the same 48 hours.
Then do the ordinary things. Move your inbound cutoff earlier than the arithmetic says you need. Ask your 3PL directly what receiving looks like on a double-container day. Split a large PO across two sailings when the SKU mix allows it, since two partly late shipments beat one fully late one. And set a date now for when you stop waiting on a container and reorder domestically, if that option exists for the product.
Our position: the risk this quarter sits in the gap between a promised arrival date and your first heavy sales day. A lot of brands left that gap too thin, because for two years they did not need it.
Receiving is where all of this lands, and it is the part of fulfillment that gets looked at least. We receive at our Canal Winchester warehouse outside Columbus and process orders within 24 hours. Location does work on the outbound side too, since 90% of customers are reachable in two days by ground and 60% of the US and Canada population sits within 600 miles. Late inventory costs less when the shipping clock after receiving is short. How we handle inbound and storage is covered on our warehouse and storage page, and the cross-border side sits under international shipping. The domestic fee picture for the same quarter is in the 2026 peak fee stack.
Want a read on whether your inbound plan has enough slack before December? Get in touch.
FAQ
What does schedule reliability actually measure?
The share of vessel arrivals that land inside the carrier's published schedule window. Sea-Intelligence tracks it across 34 trade lanes and more than 60 container carriers. July 2026 came in at 56.4%, the lowest reading of the year.
Does a cheaper ocean rate help if the ship is late?
Not for a seasonal SKU. A container that lands after your selling window closes costs you the freight, then costs you again in January markdowns. Rate savings rarely cover that gap.
How much buffer should I add for holiday inbound?
Start from your own arrival variance rather than an industry average. Pull the originally booked ETA against actual availability for your last several containers. The current average delay for late vessels is 6.06 days, so treat that as a floor rather than a ceiling.
Which carriers were most reliable in July?
Maersk led at 73.7% and was the only carrier above 70%. Hapag-Lloyd followed at 69.3%. Those two were the only names above 60%. The bottom of the table was Wan Hai at 29.8%.