The whole machine on one page — the six SCOR processes, the cost–speed–resilience trade-off, and the 1961 effect that turns a small wobble into a violent swing.
August 2026 · Case 1 of 6
As you read — hold this question
If a supply chain is just a machine with five moving parts, why does a tiny nudge at one end come out as a violent swing at the other — and can software actually fix that?
10→50
A shopper buys 10 extra boxes of cereal on a quiet Tuesday. By the time the ripple reaches the factory, they gear up to make 50.
Nobody panicked. Every link did the sensible thing — a store manager padding an order “just in case,” a distributor grabbing a bulk rate. And the system still cracked the whip. That amplification is the oldest, deepest problem in supply chains — and it is exactly where AI is supposed to help. This page is the machine, and the whip.
The standard picture
The textbook versionQuiz: SCOR & the trade-off
The machine: five links, one loop, one governor
Every supply chain maps onto the same six standard processes — the SCOR model. Five run in sequence: Plan (what people will want) → Source (buy the materials) → Make (build it) → Deliver (ship it) → Return (handle what comes back). And one wraps around all of them: Enable — data, business rules, risk, compliance. Enable is the unshowy sixth process, and it is exactly where AI governance lives; forget it in an exam and you drop marks.
The five links
Plan · Source · Make · Deliver · Return — run in order, each feeding the next.
The trade-off triangle
Every decision trades cost vs speed vs resilience. You cannot max all three — cheaper-and-faster usually means more fragile. Most AI promises are a bid to bend this triangle.
Carry this
Name all six SCOR processes — and don’t forget Enable. It’s the governance layer, and the easy mark to lose.
The oldest problem — and who saw it first
The rest of the story
The bullwhip effect: a wobble becomes a swing
The amplification in the hook has a name: the bullwhip effect. A small change in real demand grows larger at every step upstream, as each link reacts to the padded order in front of it rather than the true signal. It was first described by Jay Forrester at MIT in 1961 — decades before anyone said “AI” — and formally named and modelled by Lee, Padmanabhan & Whang in 1997, who traced it to four causes: demand-signal processing, order batching, price fluctuation, and shortage gaming.
What AI changes is not the arithmetic — it’s the eyesight. A shared machine-learning forecast can hand every link the same true demand signal from the shelf, instead of each guessing from the next one’s guess. When everyone reads the real number, the whip loses most of its crack.
The other side — does the textbook hold up?
Two catches the set materials skip. First, the classic linear SCOR chain is dated — its own custodian (ASCM) rebuilt it into the 7-process “SCOR Digital Standard” to describe multi-directional, digital chains. Second, AI smooths but does not eliminate the bullwhip (Weisz, Herold & Kummer, 2023) — and naive auto-reorder bots that all react to the same signal can actually amplify it. The barrier was never really the maths; it’s data you don’t own and trust you haven’t built.
Quick recall — without looking back
Test yourself on this case
Question 1 of 3
Name the six SCOR processes, and say which one is the governance layer.
Plan, Source, Make, Deliver, Return — run in sequence — plus Enable, which wraps around all of them (data, rules, risk, compliance). Enable is where AI governance lives.
Question 2 of 3
What is the bullwhip effect, who first described it and when, and what does AI actually change about it?
A small change in real demand amplifies at each step upstream. First described by Jay Forrester (MIT, 1961); named and modelled by Lee, Padmanabhan & Whang (1997). AI changes the eyesight — a shared forecast gives every link the same true demand signal — but only smooths the whip; it doesn’t remove it, and clumsy automation can worsen it.
Question 3 of 3
Name the three-way trade-off every supply-chain decision makes, and why AI can’t simply win all three.
Cost vs. speed vs. resilience. You can’t max all three at once — cheaper-and-faster usually buys fragility. AI’s real pitch is to bend the triangle a little (buy some resilience without paying full price elsewhere), not to abolish it.
ASCM — SCOR Digital Standard (7 processes). Weisz, Herold & Kummer, “Revisiting the bullwhip effect: how can AI smoothen the bullwhip?” Int. J. Logistics Management 34(7) (2023).
Innovare Study
Long-form video: What AI Actually Does in Supply Chains. 2026. youtu.be/HlD8PWRNDiM