A scenario

What happens if we move manufacturing

Unit cost falls on a spreadsheet and quality wobbles in the real world. The saving is permanent and starts on the day of the move. The disruption is temporary and is the only part your customers will remember.

Run this on your own numbersAll scenarios

Ships as a scenario: one lever, move_manufacturing at minus 18 percent, landing in month two, with five months of disruption and a quality dip of 0.12.

640 runs of one scenario

Mechanics

What the model does with it

One lever, four effects, three different clocks. The cost is permanent, the capital is a single month, the quality dip has an end date and the customer memory of it does not.

Month 2

Unit cost moves against a wider share than a supplier change does

The supplier index is multiplied by one plus the percentage times your supplier share of cost of goods plus 0.2. The extra 0.2 is the model saying that moving where you make something touches more of the cost base than renegotiating with one vendor does. On the invented Harborline record, with a supplier share around 0.68, an 18 percent unit cost reduction takes about 15.8 percent off the index and roughly 129,000 dollars a month off cost of goods.

Month 2

The capital cost leaves as cash in one month

The capex goes into the one off line, which is subtracted from cash that month and then reset. It defaults to 80 percent of one month of revenue, about 949,000 dollars on the Harborline example against 1.85 million of cash. Set it from the actual project rather than the default, because it is the number most likely to decide whether the run survives.

Month 2

The quality dip goes on and a counter starts

A quality drag of 0.12 is added and a disruption counter is set to five. The counter ticks down once in the month it is set, so the dip sits on the month of the move and the three months after it, and comes off before quality is computed in the fifth month of the window.

Months 2 to 5

Quality slides toward the dip rather than jumping to it

Quality moves 22 percent of the way to its target each month. Over a four month window that expresses about 63 percent of the dip at the worst point, which is a modelling detail worth knowing: the engine is gentler about a short disruption than a real factory move usually is. Widen the dip or lengthen the window if you have evidence for it.

Months 3 to 8

Satisfaction trails quality and churn trails satisfaction

Satisfaction moves 18 percent of the way to its target each month, one step behind quality. Churn carries a term in the change in satisfaction and a term in relative quality. So the customer consequence of a disruption that ends in month six is still arriving in month eight, which is exactly the pattern people misread as the move having worked.

Month 6

The drag comes off and the saving stays

The quality drag is removed when the counter reaches zero. The cost reduction has no end date. From here the run is a cheaper company with a satisfaction number still recovering, which is the shape this decision has when it goes well.

Months 4 to 12

Named accounts arrive at renewal having lived through it

Each named account has a renewal clock drawn between 40 and 130 percent of its term. The ones whose clock lands inside the disruption window weigh relative quality at 0.9 in their decision. A move that is invisible to most of your book can still cost you the two accounts that happened to renew in the wrong quarter.

The ecosystem

Which agents move, and why

The interesting thing about this scenario is that the agent who benefits and the agents who pay are on different clocks and report to different people.

  • The finance lead weights protecting margin at 0.45 and gets the benefit immediately, in the month of the move
  • The operations lead weights delivering what was sold at 0.5 and holding cost per unit at 0.3, which is the objective conflict this whole decision is made of
  • Customers do not see a cost line. They see relative quality against the strength weighted competitor quality, and they act on it at their own renewal
  • Named accounts whose renewal falls in the disruption window are the concentrated version of that risk
  • Employees and departments weighted toward delivery at 0.75 carry the operational load, and a bad enough utilisation month reaches morale and then attrition
  • Competitors do not react to your move at all. Nothing in the engine lets them use your disruption as an opening, which is optimistic
CusCustomersComCompetitorsSupSuppliersSalSalespeopleExeExecutivesEmpEmployeesInvInvestorsRegRegulatorsParPartnersYou

An invented example

The shape of the decision on Harborline Components

15.8%
Off the supplier index
An 18 percent unit cost move across a 0.68 supplier share plus 0.2
129k
Dollars a month off cost of goods
On 14.24 million dollars of annual revenue at 31 percent gross margin. Harborline is invented
949k
Dollars of capital, in one month
The default of 80 percent of monthly revenue, against 1.85 million of cash
7
Months to pay the capital back on cost alone
Before anything the disruption costs you in churn

Honestly

Where this is weakest

This scenario compresses a two year programme into one lever with four numbers on it. That is useful for framing the decision and useless for running it.

  • The saving starts at full size on day one. There is no yield curve, no learning, no scrap rate that starts high and falls
  • The disruption is a single flat drag with a fixed end date, rather than something that gets worse before it gets better
  • Supply lead time is in the ledger and the engine does not read it, so a move that doubles your lead time looks identical to one that does not
  • No freight, no duty, no currency, no tariff. For a physical product moving across a border those are frequently the whole answer
  • No dual running. Most real moves pay for both sites for a period and this lever does not
  • Competitors do not exploit the window, and the customers who leave during it can be won back in the model as easily as any other new logo

What this cannot tell you

The band, not the line

The same move, eighty times

month 0month 12best tenthworst tenth
Each line is one replication with its own draws for the disruption size, the cost move and everything downstream of them. On a manufacturing move the spread in the second half of the year is almost entirely about which named accounts renewed during the disruption window, which is why the event list is more useful than the chart on this particular scenario.

What people ask about this one

How long should I set the disruption?

Longer than the project plan says. Five months is the default and it is generous for a factory move. Run it at five, eight and twelve and see at which point the decision changes sign. If it never changes sign, you have a robust decision. If it changes at seven, you have a project management problem rather than a strategy one.

Why is the cost applied to the supplier share plus 0.2?

Because moving where something is made touches labour, overhead and logistics as well as purchased materials, and a supplier price change does not. It is a deliberate widening and it is arguable. If you disagree, enter a smaller percentage to compensate and write down why in the brief.

Can I model moving only part of the production?

Scale the percentage down by the share you are moving. Eighteen percent on half your volume is closer to nine percent entered. The quality dip should come down too, though not proportionally, because a partial move usually disrupts scheduling for everything rather than for half of it.

What about switching supplier instead of moving production?

That is a separate lever with a narrower cost base, a shorter default disruption of three months and a smaller default quality dip of 0.08. Run both against the same baseline. The comparison is usually more informative than either run on its own.

Price the disruption, not just the saving

Most cases for moving production are built on the cost line alone. This one puts the quality dip, the capital and the renewals that fall inside the window on the same page.

Build a twinAll fifteen scenarios