By Alex Stark
Not More, Better Placed: A Peak Season That Refused to End, Nike’s Exit From the S&P 100, Target Shipping the AI, and What Grocery Catering Figured Out
Labor Day is in the books, and it feels like the wheel keeps spinning faster toward the end of the year. Blink, and it will be Thanksgiving, followed ever so closely by the full-court-press of the holiday season. In my corner of the supply chain, Q4 is shaping up to be another fun-filled adventure.
Four stories this week, and three of them come down to the same idea. The win isn’t more. It’s better placed. Grocery catering isn’t growing menus; it’s making the pieces configurable. Target didn’t build a chatbot; it placed AI strategically at four specific moments. And Nike’s long slide traces back in part to a decision about where its shoes could be found.
The fourth story is about volume moving around the calendar, and it includes a correction to something I wrote in July.
Let me know what you’re seeing.
1. Peak Season Isn’t Over, and My July Math Was Incomplete
The Global Port Tracker from NRF and Hackett Associates now forecasts September at 2.31 million TEU, up 9.6% year over year, which would make it the busiest month of 2026. As recently as August, May’s 2.24 million looked like the year’s high point. July came in at 2.3 million, and August is projected at 2.29 million.
Jonathan Gold at NRF put it bluntly:
We thought the peak season would be mostly behind us by now, but that’s not the case.”
Where I Got It Wrong
On July 3rd, I wrote about tariff front-loading and told you to plan for a genuinely soft Q4 because a big chunk of the fall’s volume was already on the water. On July 17th, I mentioned it again. The front-loading logic was right. However, I was wrong about the magnitude. Volume did get pulled forward, but consumer demand held up well enough that Q3 did not drop off as the borrowing-from-the-future model predicted. Ben Hackett noted that consumers might reasonably have been expected to turn cautious as cost pressures persisted, but they didn’t.
Here’s the number that reframes the entire year. Full-year 2026 is forecast at 25.7 million TEU, up just 1% from 2025’s 25.4 million. The first half was up 1.1%. So, the annual total barely moved. What changed was the distribution. Tariff deadlines pulled cargo around the calendar, and the traditional late-summer peak reasserted itself anyway. Same volume, different shape.
Looking ahead, October is forecast at roughly 2.11 million, November at 2.0 million, December at 2.03 million, and January 2027 at 2.09 million. Normalizing, but from a higher base than last year. Hackett also flagged reports of vessel delays and longer cargo transit times, which lines up with the Panama Canal transit cuts I wrote about two weeks ago.
The lesson I’m taking from my own miss… is that front-loading models are good at predicting where volume moves. They are much worse at predicting whether the underlying demand holds. I conflated those two things in July, and it’s worth flagging because much of this fall’s capacity planning was built on the same assumption.
2. Nike Is Out of the S&P 100 After 18 Years
S&P Dow Jones Indices announced on September 4 that Nike will be removed from the S&P 100 before trading opens on September 21, after joining the index in December 2008. Simon Property Group, Honeywell Aerospace, and Colgate-Palmolive will also be removed on the same date. All of them will remain in the broader S&P 500.
The scale of the decline is hard to overstate. Nike’s stock is down roughly 78% from its November 2021 peak, with about $224 billion in market value gone. Market cap has fallen from around $264 billion to roughly $57 billion, including a 36% drop in 2026 alone. Shares closed at $38.40 on September 4, near a twelve-year low.
The Replacements Tell the Real Story
Dell Technologies, Palo Alto Networks, Arista Networks, and SanDisk are moving in. All four are in information technology. Arista makes the switches that connect AI data centers. SanDisk makes flash memory. In the filing itself, Palo Alto is paired directly against Nike’s deletion, and Arista against Simon Property Group. Four consumer and property companies out, four pieces of AI infrastructure in. The market is quietly redefining what counts as essential.
The Part That Belongs in This Week’s Theme
Nike’s slide traces largely to a 2020 decision by former CEO John Donahoe to exit wholesale channels and push direct-to-consumer (DTC). That was a distribution strategy, not a product one. Nike traded placement for control and margin, and the brand became harder to find. Wells Fargo’s latest note lists Nike among the largest laggards, with social media mentions down 10% in the second quarter. Current CEO Elliott Hill has been rebuilding wholesale relationships since late 2024. The company’s performance and loss of value are trying to get back on track, but the turnaround is taking longer than investors’ patience allows.
This fascinating article gets into the details. A former Nike executive wrote it, having witnessed the company’s steady, even rational (at the time) decisions.
Back in June, I wrote about Duluth Trading handing its fulfillment to Amazon, and the reasoning was simply that customers were already there. Nike made the opposite bet six years earlier, pulling back from where customers already were. Both are channel decisions, and both are reversible. Nike’s decision just cost $224 billion to learn.
Being available where people already shop is worth more than owning the transaction. That’s an uncomfortable lesson for anyone building a direct channel today.
3. Target Shipped the AI, and It Isn’t a Chatbot
Target announced a suite of AI shopping features this week. Review Insights launched in June, using AI to summarize product reviews so shoppers can skip reading through them. Photo Search launched in August, letting app users upload or snap a photo to find similar products without describing them. Buy Again surfaces frequently purchased items and relevant deals. Continue Shopping picks up where a session left off.
The design decision is the whole story. Target’s approach differs from the chat-first assistants launching everywhere else. Rather than one conversational tool sitting at the front door, AI is embedded at distinct points in the shopping journey. Finding a product. Deciding on it. Buying it again. Sarah Travis, EVP and chief digital and revenue officer, said the goal is to help guests find what they need faster and shop with confidence.
Target’s ongoing positive momentum in 2026 suggests that the turnaround under its new CEO is gaining traction. Looking at the retail landscape, they are focused on long-term recovery to increase sales.
Why That’s the Right Read
Three weeks ago, I wrote about the cost of AI failures. More than a third of consumers report immediate frustration the moment they learn that support is AI-powered, and three in five repeat themselves exactly once to an automated system before abandoning it entirely. Given those numbers, making one big chatbot your front door is a genuine risk. Target, instead, employed AI and personalization at four moments where it removes a specific annoyance. Review Insights has already driven conversion and add-to-cart activity, the company says.
There’s a three-week arc here worth mentioning. On August 14th, Target hired its first chief AI officer and paired the announcement with a UX promotion, and I argued at the time that the pairing was the point. On August 28th, the company named consistency and in-stock reliability its theme for the year. Now it has launched features that remove friction at specific decision points. Target’s goal was deliberate. It named an owner, defined the outcome, and delivered.
That sequence is exactly what Gartner said was missing when it found that two-thirds of supply chain digital spend flowed to AI, while 55% of executives couldn’t say what any of it returned. The useful question was never how much AI you deployed. It’s how many specific annoyances you removed. Those two questions produce very different roadmaps.
4. Grocery Catering Says the Trend Isn’t Bigger Menus
A Supermarket Perimeter piece on how grocery catering is changing has the week’s best quote. Cori Nobriga, Director of Service Deli at The Save Mart Companies, on what customers are actually asking for:
The trend isn’t bigger menus. It’s more intentional entertaining.”
Customers are balancing health goals, dietary requirements, protein-forward eating, family traditions, and convenience all at once. They want menus that reflect the specific people they’re serving. Another straightforward quote from Nobriga states that catering has evolved from feeding a crowd to creating an experience. This specific article is grocery-based, but the implications apply to any segment.
The Answer Is Components, Not SKUs
Here’s what matters for anyone running an operation. Grocers aren’t answering this demand with more finished products. They’re answering it with configurable components.
- Save Mart’s “Perfect Bite” bundles combine complementary meat, cheese, produce, and crunchy elements in a single purchase, with a QR code showing how to assemble them: ingredients plus the confidence to use them.
- Giant Food is seeing demand for build-your-own concepts and restaurant-inspired presentation.
- Wegmans offers customizable Asian, Latin, and Mediterranean bowls, gluten-free options, Meals2Go, and recently launched hot catering.
- Hy-Vee leans on potato bars, taco bars, and burger bars, which let guests build their own plates from a shared component set at an economical price point.
Last week, I wrote about Under Armour cutting SKUs by 25% and BJ’s targeting a 20% reduction, and now here’s another category adding customization. Those aren’t opposites. A taco bar is one component set that produces a hundred different plates. Fewer inputs, more outcomes. That’s the version of variety that doesn’t punish your warehouse, and it’s the answer to BJ’s problem of cutting SKUs and cutting sales right along with them.
The occasions are shifting, too. Abby Despins, Vice President of Education and Industry Relations at the International Dairy, Deli, and Bakery Association, notes that catering is no longer reserved for weddings and major corporate events. It’s now used for family gatherings, workplace meetings, sporting events, and casual entertaining. Smaller, more frequent. Younger consumers, in particular, want convenience, personalization, and online ordering.
And here’s the detail most coverage might overlook. Hy-Vee’s Bryan Polc (Assistant VP of Food Service) mentioned fulfilling orders with specific delivery requirements, such as dropping off food at multiple locations or serving the same meal to one group at different times. That isn’t a menu problem. That’s routing, timing, and temperature control, executed against a promise made at the point of sale.
Mass customization only works if the fulfillment behind it is disciplined. Configurable products create more variability downstream, not less, and somebody has to absorb that in picking, staging, and delivery windows. That absorbing is the work.
Not More. Better Placed.
Full-year imports are up 1%, so the volume didn’t grow; it shifted. Nike has plenty of product and lost the places where people were already shopping. Target skipped the chatbot and deployed AI at four specific moments. And grocery catering figured out that customers want the pieces to fit the occasion, not a longer menu.
When things get challenging, the instinct might be to add. More volume, more features, more SKUs, more control. Every one of these stories points the other way. Fewer things, placed where they do the work.
That’s harder than adding because it requires knowing which parts matter. Most of us learn that by getting it wrong first, which is what the first story in this post was about.
Bonus #1: Salem for Spooky Season
Halloween is right around the corner. We’ve already got the mums around the house. Next up are the pumpkins. Really into the spooky season? Travel + Leisure’s travel pros picked Salem, Massachusetts, as the best spot to soak in all the creepy, scary, haunted fun. The town, famously (or infamously) known for the 1692 witch trials, leans in and owns it.
I speak from experience. It’s a treat.
Bonus #2: Internet History Roulette
Every once in a while, it’s useful for my research to check what a website looked like. The Wayback Machine is a digital archive of the web, and in a tribute to the site, I stumbled upon this cool little interactive called Internet History Roulette. Give it a spin to see what things on the World Wide Web looked like years ago.
You’re welcome.
Bonus #3: The Happiest States in America
Readers of this blog know that I’m a sucker for a map. Any map, really. But especially ones that show my beloved Pennsylvania in a better light, better ranking, better anything compared to Ohio. Here’s another one: The Happiest States in America, as compiled by WalletHub.
One More Thing
I would be remiss not to mention the countless acts of courage and sacrifice by the first responders on that awful Tuesday morning twenty-five years ago. Their bravery in the face of extreme danger is among the purest expressions of humanity.
To all those affected by that day and its aftermath, please know that we will never forget.
It costs nothing to be kind.
