The cookie test

Chicago Booth published a piece in May asking four supply-chain people a plain question: what should companies actually do about all this uncertainty?

Four good answers. Also four answers written for people who already speak supply chain.

So I rewrote them as a bake sale.

Here's the test I use, and it has never failed me: if you can explain a risk to an eight-year-old, you understand it. If you can't, you are hiding behind vocabulary. Boards do this. Consultants do this. I have done this.

The setup. You're making 200 cookies for the school bake sale on Saturday. You need flour, butter, sugar, and chocolate chips. You buy all of it from the grocery store down the street.

That's a supply chain. Everything below is the same thing with more zeros.

Pushkar Deshpande, who runs product at the logistics platform Turvo, says to "build out your business operations to account for potential expenses adjusted for outside factors and disruptions."

Translation: most companies build the plan at today's price and treat every increase as a surprise. It isn't a surprise. It's Tuesday.

The number that makes this concrete: the average U.S. statutory tariff rate sat at 11.0% on August 24, 2026, and the Yale Budget Lab projects 11.8% by year-end under current law. If your landed-cost model still runs on 2024 duty rates, your margin plan is a work of fiction.

Deshpande's second idea is a digital twin, a software copy of your supply chain you can break on purpose. Do the bake sale on paper on Wednesday. Find out you're short on butter while there's still time to buy butter.

Your store is a tier-one supplier. The farm behind it is tier two. The farm's fertilizer supplier is tier three. That's where the risk actually lives, and almost nobody can see it.

McKinsey's 2025 survey of 100 supply-chain leaders found that most companies still understand their risk only at tier one, and that tier-two and tier-three visibility got worse in 2023 and 2024 than it had been in 2022. It went backwards. Companies were cutting digitization budgets under cost pressure at exactly the moment the map got more complicated.

Levi DeValve, who teaches operations at Booth, argues for shortening the chain: onshoring and nearshoring, and paying more per unit on purpose. "The more you can build in flexibility, adaptability, and the capacity to quickly pivot, the more resilient your supply-chain network will be." A shorter chain has fewer places to be blind.

What it's worth: one manufacturer McKinsey worked with found 10 to 20% of its cost of goods sold was sitting in at-risk supply, and 5 to 15% of that exposure could be fixed by restructuring the network. That's not a resilience project. That's a margin project wearing a resilience jacket.

Melissa Hadhazy at Ingram Micro makes the least exciting and most useful point in the article: "Rather than focusing on the next big thing, you're better off looking at what you've already done with your supply chain and making sure that you're doing it correctly."

Most disruption is not geopolitical. It's a handoff nobody owns, two teams working from different spreadsheets, and a process that was designed for a volume you outgrew two years ago. Hadhazy's answer is people and change management before technology, which is unfashionable and correct. The frontline engineer who knows why the line actually stops is not replaceable by a dashboard.

Before you buy the ninth tool, find out how much of the third one you're using.

Sundeep Khera, global chief underwriting officer at AXA XL, puts it this way: it's "imperative for companies to embrace digital transformation and upgrade to smart tools that can help them better track the provenance of goods, parts, and components."

This is the part I spend my days on, so forgive the enthusiasm.

The teacher is the regulator. The teacher is also your biggest customer's procurement team. RoHS, REACH, EUDR, conflict minerals, PFAS reporting: the question underneath every one of them has the same shape. What's in it, where did it come from, prove it. A product you can't document is a product you can't ship. Nobody bans the cookie. It just quietly never makes it onto the table, and you find out at the border.

That's four Booth experts and about a thousand words of magazine.

Here's why I write this newsletter the way I do. Most operators I talk to already know all four of these. What they don't have is a way to say them out loud to a board director, a plant manager, and a 22-year-old analyst in the same meeting and have all three walk out with the same picture in their head. Simple language does more than make things friendly. It's what lets a decision travel from the boardroom to the plant floor without changing shape on the way down.

If your team can't pass the cookie test on your own supply chain this week, that's the finding.

— Kunal