The Email That Stops a Factory
06:40 on a Tuesday: your only supplier of a critical part says twelve weeks. Here is the arithmetic that turns those words into 440 unbuilt ovens, and the eleven desks it is about to cross.
What you'll learn
- Turn a supplier's delay into the only number that matters — units you cannot build
- Meet the eleven people a shortage travels through, and the question each one owns
- Understand why a shock behaves differently from a project
One email, eleven desks, and a shortage that has to be shared out among real customers by Friday.
At 06:40 on Tuesday 3 March, an email arrives at Calder Thermal from a supplier in Penang. It is four sentences long, perfectly polite, and it will consume the next twelve weeks of about forty people’s working lives. Vantor Electronics, the only company that makes the CB-40 control board, has had its microcontroller allocation cut by the chip maker. CB-40 deliveries will be twelve weeks late.
Not cancelled. Late. That distinction matters more than it sounds, and by the end of this course you will see why it changes almost every decision in the story.
What Calder Thermal actually does
Calder Thermal makes commercial ovens in Rockford: $50m of revenue a year, 450 staff, one factory. There are three lines — the Meridian combi oven for hotels and restaurants at $14,000, the Compact for in-store and independent bakeries at $6,000, and the Rack for high-volume industrial bakeries at $22,000. Eighty ovens leave the building in a normal week, which is about $960,000 of revenue at a blended gross margin of 33%.
Every one of those eighty ovens, whichever line it comes from, contains one CB-40 control board. That is the whole problem in a sentence. A component shared across every product is efficient to buy, cheap to stock, simple to service — and it means a single supplier can stop the entire factory rather than a third of it.
The shape of this course
Case one in this collection followed a slow decline that somebody had to go looking for. This one starts with a shock nobody can ignore. Same skeleton — what lands on a desk, what that person adds, what they hand on — but the clock is running, and most of these decisions cannot be undone once taken.The arithmetic that turns an email into a problem
Here is the sum that every one of the eleven people in this story works from, and it is worth doing slowly because everything downstream is an argument about these numbers.
Calder uses 80 boards a week, one per oven. There are 520 CB-40 boards in the stores in Rockford. That is 6.5 weeks of full production sitting on a shelf. The delay is twelve weeks. So for 5.5 weeks the line has no boards at all, and at eighty a week that is 440 ovens that cannot be built.
Against those 440 missing ovens sit 960 firm orders for the twelve-week window — real customers with real delivery dates, some of them contractual. At an average selling price of about $12,000, the exposure is roughly $5.3m of revenue and $1.7m of gross profit. Calder’s whole quarter is $12.5m, so this is not a rounding error; it is a large fraction of a quarter, and the quarter closes in week nine.
The words this crisis is conducted in
- Sole source
- Only one supplier is qualified to provide a part. Cheaper and simpler right up until the moment it isn’t.
- Lead time
- How long between ordering something and being able to use it. Twelve weeks late means every promise built on the old lead time is now wrong.
- Weeks of cover
- Stock on hand divided by the rate you consume it. 520 boards at 80 a week is 6.5 weeks of cover — the length of the fuse.
- Allocation
- Deciding who gets a scarce item and who does not. In a shortage this is the single most consequential activity in the company.
- Liquidated damages
- Penalties written into a contract for late delivery, agreed in advance so nobody has to argue about the cost of lateness afterwards.
Four levers, none of them free
By Tuesday lunchtime the options are already on a whiteboard, and every one of them costs something real.
Expedite. When Vantor resumes, air-freight the batches instead of shipping them. That claws back about two weeks and costs $85,000 — real money for time, which is the most common trade in operations.
Find another supplier. Kestrel Micro in Poland can build a compatible board, but it takes ten weeks to qualify a new electronics supplier for a product that carries safety approvals, plus $180,000 of tooling, and the boards cost $460 each against Vantor’s $320. Ten weeks is most of the crisis, which is the uncomfortable lesson: the alternative supplier you set up during an emergency mostly helps with the emergency after this one.
Buy on the open market. Brokers have CB-40 boards today at around $1,100 each, three times list, with provenance nobody can verify and a genuine risk of counterfeit parts inside a product that heats food in someone’s kitchen.
Redesign around the shortage. Engineering can respin the board to use a microcontroller that is actually available — in fourteen weeks, including safety recertification. Longer than the disruption it is meant to solve.
The trap in every shortage
Each lever looks like a solution on its own and none of them is sufficient alone. The answer at Calder will be a combination, chosen by different people on different days, and the combination is what makes this hard to manage — because nobody sees the whole of it except the person at the end adding up what it cost.The software this shock travels through
The ERP holds what is true; the planning tool decides what happens next; the CRM holds who it happens to.
In a manufacturer the ERP is the spine — SAP or an equivalent — because it holds the one version of stock, orders, costs and lead times that everyone must argue from. When Hannah says 520 boards, she is quoting the ERP, and the reason nobody challenges the number is that challenging it means challenging the system the accounts are built on.
Around that spine sit three others. A supply planning tool models what happens if a date moves, which is the only way to see a twelve-week delay ripple through eight hundred order lines without doing it by hand. The CRM holds the commercial truth the ERP cannot: which customer is strategic, which is dual-sourcing, which has a penalty clause. And a reporting layer turns all of it into the handful of numbers the board actually watches. Notice the pattern from the first case in this collection repeating: the decisive systems are the ones people already work in every day.
The four systems this crisis runs on
- SAP (the ERP)
- Enterprise resource planning: stock, orders, bills of material, costs and lead times. The single official version of what the company has and owes.
- Kinaxis / SAP IBP
- Supply planning software. Models what happens to every downstream date when one supply date moves — the difference between a plan and a guess.
- Salesforce (the CRM)
- Holds the commercial context the ERP has no field for: contract terms, relationship history, and which accounts are quietly shopping elsewhere.
- Power BI
- The reporting layer. Turns the crisis into the six numbers the board tracks daily.
The eleven desks, and the question each one owns
Where this email is going
- Ravi Menon, COO
- What are we actually optimising? Revenue, cash, or relationships — because you cannot protect all three, and every desk downstream needs to know which one wins.
- Hannah Wolcott, Procurement
- Can we get boards anywhere else? Runs the alternatives to ground, including the ones that look attractive and are not.
- Dev Anand, Engineering
- Can we design around it? Explains why the obvious fix takes fourteen weeks, and starts the one that prevents the next crisis.
- Tomasz Nowak, Supply Planner
- Who gets the 520 boards we have? Turns a shortage into an allocation, which is where the pain gets distributed.
- Grace Okonkwo, Manufacturing
- What does that plan cost on the floor? Resequencing a line is never free — changeovers, batch sizes and idle people all have prices.
- Yusuf Rahman, Demand Analyst
- Which of these orders are real? Separates firm demand from forecast, and finds the double-counting that would otherwise send scarce boards to phantom customers.
- Claire Beaumont, Finance
- What does each option cost, and when? Prices expedite against penalties, and separates revenue lost from revenue merely moved.
- Marcus Reid, Sales
- Which customers do we protect? Decides where scarcity lands, using strategy rather than order size.
- Siân Pritchard, Customer Service
- How do we tell them? Turns a decision into four hundred conversations, and controls whether customers stay angry.
- Ingrid Sørensen, Risk
- Why could one supplier do this to us? Asks the question that should have been asked three years ago, and makes it stick this time.
- Eleanor Vance, CEO and the board
- What changes permanently? Weighs what the shock cost against what it revealed.
Why a shock is not a project
Three differences shape every module that follows.
The first is that the clock is not yours. In a project you choose the pace and negotiate the deadline. Here the deadline was set by a chip allocation decision made by strangers, and it does not care about your quarter, your holidays or your board meeting.
The second is that decisions bind quickly and cost money to unwind. Book the air freight and the $85,000 is spent whether or not it turns out to be needed. Tell Nordfoods their ovens are late and you cannot untell them. Case one had a project team choosing thresholds they could tune later; nobody here gets to iterate.
The third, and the one people find hardest, is that doing nothing is also a decision, and it is usually the worst one. Scarcity allocates itself if you let it: the loudest customer, the most senior account manager, or simple first-come-first-served will decide who gets an oven. All three of those produce a worse outcome than a deliberate choice, and none of them leaves anyone accountable.
The bottom line
A supplier’s delay becomes a business problem only once somebody does the arithmetic: 520 boards, 80 a week, 440 ovens that cannot be built. From there it stops being an operations issue and becomes, in order, a planning problem, a money problem and a customer problem — and the four available levers all cost time, cash or trust.Whose problem is it?
Read each situation and decide which desk owns it, then tap a card to check.
Quick check
1. Calder holds 520 boards and uses 80 a week against a 12-week delay. How many ovens cannot be built?
2. Why doesn't qualifying Kestrel Micro solve this crisis?
3. What makes a single shared component both efficient and dangerous?