Why Lead Times Slip

This one is about us. One thing we finally understood, and four we still haven’t fixed. There is not a single statistic in here, because we never measured it properly. More on that at the end.

1 · Lead time is the service

Naval once said something like: where your attention goes, your life goes.

A shop works the same way. Where the owner looks, that is where the shop goes.

At Eternal CNC, our attention has always sat on two things: product and service. We believed those were the only two things that actually deliver value to a customer.

Sounds like a slogan. But those two words are expensive to mean.

Making them real takes an entire team pulling the same direction, and that costs money:

Some of it shows up on the books:

• A part comes out wrong. Under the old habit, it could still ship — “close enough.” We scrap it instead. That is cash you can watch disappear in one second.
• An old machine still turns. Replacing it makes no sense on paper. We replaced it anyway.

Some of it never shows up on the books:

• We leave more stock on the material than we need to. The extra gets booked as “consumption,” not as “cost.” But a light cut can be taken again. A heavy one cannot.
• The tolerance on the drawing, the customer may never put calipers on. We hold it anyway. The customer not checking does not give us permission to skip it.
• Good tooling costs more than cheap tooling. Per part, the difference disappears into the noise. We still buy good tooling.
• A part is awkward to hold. The fixture gets rebuilt more than once before it repeats. None of that fixture work ever goes on the quote — workholding rarely has a line item.
• First article, we make several and keep the best one, rather than make one and ship it. The extras are scrap.
• On a few quotations we did not come in lowest, and we lost the job. Looking back, those prices would have been impossible to produce at.
• We would rather leave a position empty than hire the wrong person. An empty station is lost capacity, and lost capacity tightens lead times. We accept that trade.

And some of it is just friction:

• A subcontractor who doesn’t take quality seriously. Argue it back and forth, and you spend goodwill you can’t get back.

We have taken a lot of hits for those two words.

The direction still holds. We believe doing the right thing pays — just slowly.

But here is the part we got wrong for a long time.

We thought “service” meant attitude. Fast replies. How you handle things when they go wrong. We did not think of lead time as service at all — and it is the hardest part of it.

A customer will not forgive a late delivery because you were pleasant about it.

A part can be beautiful. If it arrives three months late, those three months — the waiting, the line the customer could not start, the promise they could not make to their own customer — that is service we failed to deliver.

Getting a customer’s problem handled, so they have one less thing to worry about: that is a high form of service. It does not appear on a part number or a quotation. But customers weigh it every single day.

So our attention was not in the wrong place.

Our understanding of the word “service” was missing a piece.

The piece we were missing was time.

2 · Subcontractors: not strong enough, not enough of them

This one is harder than understanding it.

When capacity runs out, you subcontract. That is normal for a small shop. But there are two shortages.

Not strong enough

Over ten-plus years, inconsistent anodizing has come up more times than we can count.

An anodizer cannot guarantee that every batch matches the sample colour. When it doesn’t, you re-anodize. If that doesn’t work, you anodize a third time.

Here is the part customers never see: every anodizing cycle moves the thread dimensions.

Nobody thinks about this. They see the colour is off. They don’t see that the thread has changed. One time, after repeated re-anodizing, an entire batch failed on thread tolerance — scrapped in full, remade from raw material.

That order didn’t ship a few days late. It started over.

Not enough of them

Another time, a subcontractor we had not worked with much promised a thousand pieces inside the window.

A few hundred showed up. Many of them failed inspection — judging by the condition of the shipment, they likely never checked anything before it left.

We sent a large portion back. The gap was serious. That order was saved only by clearing our own machines and running flat out.

With only a handful of suppliers to choose from, you have no leverage on price and no backup when someone fails.

Being narrow costs you twice:

• No negotiating power. You take the price you’re given.
• No fallback. One supplier stumbles and the whole batch stops, while the customer waits.

And that bench of suppliers is not something you build when you need it.
It is not “we couldn’t find a good subcontractor this week.” It is we didn’t spend the years building the bench.

That one is on us.

3 · The quality system is incomplete, so the lead time was never real

This is the one that hurts most, and the one nobody talks about.

Plenty of shops do “deliver.” The schedule was built. The parts went out the door.

But a large portion of what went out does not pass inspection.

Which means the lead time you quoted was fiction.

Real lead time isn’t the day it leaves your dock. It’s the day your customer accepts it. In between sits a return, a rework, a re-schedule. Fast case, three to five days. Slow case, it feels like it will never end.

What damages the relationship isn’t the money. It’s the uncertainty.

A customer waits once. Asks twice. By the third time they stop asking — they find another supplier. And what wears people down is not the loss. It is the hanging feeling of not knowing when it will be fine.

We are the same on our side: a batch comes back, the floor has to re-slot it, every other customer’s schedule shifts, everyone starts firefighting, and firefighting makes the next mistake more likely.

A gap in the quality system shows up as a lead time failure.

4 · Material suppliers

The fourth one, easy to overlook, and it belongs near the top.

If material doesn’t arrive, nothing starts. And this link has more traps than people expect.

Grades and stock don’t line up.
Common grades sit on a shelf. Special grades — 17-4PH, Ti-6Al-4V, some imported aluminium — are ordered in. “In stock” and “ships tomorrow” are two different statements, and a week can sit between them.

Small volumes have no priority.
A small shop buys tens of kilos. A large account buys tonnes. When material is tight, guess whose order ships first.

You find out after it lands.
Wrong grade, wrong hardness, not enough stock left for the finished size. None of this is rare. By the time you find out, half the lead time is gone, and the return-and-replace is another cycle.

Payment terms versus stock on hand.
Cash up front gets you material. Terms get you an approval queue. Choose terms for cash flow, and the schedule pays for it.

The most dangerous thing about this one: it isn’t on your schedule. On the schedule, material shows as “arrived.” It hasn’t. You find out too late.

5 · All four are one thing

Four causes, but dig down and they are one:

Lead time failures are not a scheduling problem.

None of these is fixed by trying harder.
All of them cost time, money, and people to build.

Which is why lead times at a small shop rarely get fixed in three months.
Anyone telling you “we’ll be on time once we install this system” is selling a system.

6 · What we changed

Not theory. Two things. Both cost money. Both are still running.

One: surface finishing only goes to shops we trust, even at a higher price.

The old habit was to go where it was cheapest. Then we learned that the saving doesn’t cover one rework.

Now it goes to a small set of shops that do it well. Unit price is higher. Colour variance and re-anodizing dropped a lot.

No clever technique here. Just an admission: good process is worth what it costs.

Two: we deliberately leave machines idle.

This one runs against instinct.

Every small shop wants spindles running 24 hours. An idle machine is money lost. We do the opposite: we keep a few machines permanently unscheduled.

Why?

Because something always breaks. A subcontractor drops the ball, or a customer needs an urgent run. At that moment, whether you have capacity decides whether you can rescue it or only apologise.

An idle machine is a cost on the books. It is also one of the few things that can be converted into delivered parts within two days.

Both of these work the same way: pay first, see the result later.

Is it enough? Not close. But these two are what we can show today.

7 · One question for other shops

If your lead times slip too, don’t start with the schedule.

Start here: last batch you shipped — how much of it got held at inspection?

That number is more honest than the schedule.