If you are a Head of Product or a procurement lead, what is the real cost of a three-month production delay?
Not the spreadsheet version. Not the extra freight line or the extra month of spreading the tool cost. The cost that does not sit neatly in a cell.
A critical component sits on NC (non-conforming) hold on a UK (United Kingdom) goods-in floor. Six weeks, sometimes more. The assembly line has nothing honest to build. The launch date that sales have been working to moves. That is when the cheap overseas quote stops looking clever.
The costs that never make the PO (purchase order)
Competitors take the window you were aiming at. A retailer or a key account does not hold a slot because your first-article inspection is still open. A sales team that was wound up for a date loses the conversation and has to reopen it cold. Customers who were told "this quarter" start to treat the brand as late by default.
None of that appears as a unit-cost variance. It appears as a missed season, a product that slips a range review, a retailer who gives the shelf space to someone else, or a customer who builds you into their product and then dual-sources you off the next programme.
Fragmented, low-cost sourcing looks excellent in a procurement review. Many factories. Lowest landed cost. Then the variation arrives - colour, shrinkage, the start of a thread, plating thickness, a drawing revision that only one supplier received - and the UK line stops. You did not buy a cheap part. You bought a single point of failure with a long boat ride behind it.
How a delay actually starts
It rarely starts with a ship that is late. It starts earlier.
The design is signed off before anyone has agreed what "good" means on the dimensions that actually matter: which features will be measured, to what tolerance, and with what gauge or method. Without that, first-offs get argued over instead of accepted or rejected.
The tool is then cut for a different polymer grade than the one that will run in production. Shrinkage is not the same from grade to grade, even inside one family. If the cavity is compensated for the wrong material, the moulded part will not sit on the drawing once you switch to the real production polymer.
First-off samples look near enough, so T1 (first tool trial) is treated as a visit to tick a box rather than a hard pass or fail. In proper tool try-out, T1 is a hard pass or fail: either the tool and the process are good enough to move on, or they are not. If you celebrate "close enough" and keep the launch date, you have already accepted the delay. You have just not put it on the calendar yet.
The dim report (dimensional inspection report) is thin - missing balloons, missing datums - or the parts are measured in the wrong condition. Nylon is the usual example. A part measured dry, straight out of the tool, is not the same size as a part that has taken up moisture in a warehouse or in service. If you bless the dry numbers and ship, the snap-fits and bearing fits move later.
A concession is accepted "just this once". The next batch is made the same way. The exception has become the process.
Then goods-in finds what sampling should have found. An NC hold is the honest name for work that was deferred.
Six weeks with parts locked in NC is not a quality team being difficult. It is the programme paying interest on decisions that were not closed before steel was cut, or before the PO was split across three vendors who do not talk to each other.
Agree what good means before steel is cut - ballooned drawing, gauges, a dim report someone else can repeat.
Low piece price, high programme risk
A pound saved on a component is visible. A three-month slip is spread across sales, freight and reputation, so it gets discounted.
Do the uncomfortable arithmetic once, in private, before you place the next order.
What is the revenue attached to the original launch month? What does the sales team lose if they miss the customer's buying window? What does extra air freight cost when you finally clear the hold and try to recover the calendar? What does a second sampling loop cost in tool modifications, travel and engineering time? What is the cost of a line of people standing around a kit that is one bracket short?
Add brand. A hardware product that arrives late twice does not get the benefit of the doubt on the third programme. That is not sentiment. That is how procurement scores supplier risk.
What "managed from the UK" is supposed to prevent
Supply from more than one region is not the problem. Unmanaged supply from more than one region is.
The UK, Europe, China and Taiwan can all sit in one programme if someone owns the drawing, the production polymer, the inspection standard and the recovery plan. If those live in four inboxes, you will meet them again on the hold bench at goods-in.
The work that protects a date happens before the boat.
Agree the dimensions that are critical to function, and how they will be measured. "To print" is not a method. You need a ballooned drawing, agreed gauges, and a dim report that a second inspector can repeat.
Lock the production polymer grade and the cosmetic standard before T1. A first-off moulded in the wrong grade is not a first-off. The shrinkage, the fill and the surface will all change when you put the real material in.
Know which supplier is allowed to change what. A mid-stream tool modification in one plant that the assembler never sees is how you get a batch that almost fits.
Plan the hold. If a lot fails, who decides, how fast, and with what next step - sort, rework, recut steel, or run a second source. An NC hold with no owner is a delay with a polite name.
A delay is a choice you can still unmake
You cannot buy back a missed launch window. You can stop treating date risk as something that starts at goods-in.
Bring supplier capability and single-point-of-failure checks into the design review, not the recovery meeting. If one component can stop the finished box, it needs a second source or a harder inspection gate, not a hopeful lead time.
Treat T1 as a decision, not a photograph. If the samples are "almost", the date is already moving. Say so then.
Stop splitting the BOM (bill of materials) across the cheapest bidder on every line. A coherent supplier set with one accountable owner is slower to quote and faster to ship.
Piece price will still matter. So will the date you told the market. The second number is the one customers remember.