The 6-Point Quote Checklist That Stops Useless Custom Parts

When This Checklist Helped

I've managed custom manufacturing orders for seven years. In that time, I've shipped parts that worked and parts that should never have left the shop. The most expensive lesson wasn't scrap metal. It was accepting a request before understanding the problem.

In September 2022, a customer sent us a design for a replacement brake pedal arm. The note on the order said, “Why do I have to press my brake pedal so far?”

We should have stopped there. Instead, we quoted a billet aluminum part, programmed the CNC, and made something that looked better than the factory original. The customer installed it and reported no change. The original arm wasn't even the problem. A brake specialist later found the real issue in the hydraulic system. Our part was a well-made piece of metal that answered a question nobody asked.

That mistake cost the customer a delayed repair and cost us roughly $1,700 in machine time, material, and embarrassment.

So I built a checklist. It doesn't make us perfect, but it stops the most common failure I see in custom manufacturing: printing or machining a part before anyone defines what “fixed” actually means.

The 6-Point Pre-Production Checklist

1. Separate the symptom from the specification

A repair order starts with a symptom. A manufacturing order starts with a specification. If you mix those two, you end up making parts that are technically perfect and functionally useless.

The brake pedal request was a symptom: pedal travel too far. A manufacturing specification would have included the pedal ratio, pivot location, mounting dimensions, applied force, material grade, and the safety standard the part must meet. We had none of that.

Now every order has to pass one gate first: Can you write down what this part must do, in numbers?

If not, stop. Go back to the customer and ask for the real problem, or refer them to someone qualified to diagnose it.

2. Choose the process based on the function, not the machine you want to use

I like our 3D Systems powder 3D printer technology. It handles complex internal geometries and short-run metal parts that would be difficult to machine conventionally. 3D Systems 3D printers are genuinely useful in a job shop. But a printer is not a repair manual.

When the final part has to survive fatigue loading, impacts, or high heat, we now ask whether additive is the right answer or whether CNC machining or a forged blank makes more sense. For sheet metal assemblies that need a sealed joint, a welding laser service might be the right step after machining. Every process has limits.

The point is not to choose the coolest tool. The point is to make a part that works in the real environment. If we can't prove the material and process will survive that environment, we don't quote the job as if we can.

3. Simulate before you cut or print

We run every CNC program through a toolpath simulation before it touches the machine. A CNC turning center simulator catches tool holder collisions, poor workholding clearance, and bad finish passes before they become scrap.

In January of last year, I skipped that step on a rush job. The tool holder clipped a finished diameter and damaged the insert, the turret, and the part. The repair cost over $900 and added four days to the schedule. All because I didn't want to spend ten minutes in the simulator.

For metal 3D printing, I use the same logic. The build simulation shows me where supports are risky, where the part could distort, and where powder removal will be impossible. A few hours of simulation is always cheaper than a failed build or a collision.

4. Make a first article, then measure it against the original intent

A CAD model is not a part. The first real part is where tolerance, finish, and distortion problems show up.

We once machined a flange that matched the digital model perfectly. Then we tried to bolt it to the mating component. The bolt holes were clocked half a millimeter off because the datum on our fixture didn't match the datum on the customer's system. The profile was perfect. The function was wrong.

Now every job gets a first-article inspection before full production. If the part must fit another component, we verify that fit physically, not just in the software.

5. Define the acceptance test and the paper trail

A part that “looks right” is not a pass. The acceptance test should be written down before the part is made.

If the part is structural, the test might include a load test or a pressure test. If it's a cosmetic bracket, a dimensional check and visual standard might be enough. If the customer is in aerospace or defense, the documentation requirements are even stricter. In those cases, the material certificate, process records, and inspection reports are part of the deliverable. ISO 9001 and AS9100D both expect controlled production and clear documentation. We treat those standards as the baseline for critical work.

If a customer can't specify an acceptance test, we now say so clearly before quoting. That sentence alone has saved us from several embarrassing orders.

6. State limitations honestly on the quote

This is the step that sounds counterproductive but actually builds trust.

When a customer asks for a custom component that affects vehicle safety, I don't pretend our shop can validate the whole vehicle. We can make a pedal arm to print, but we cannot certify that it works with every brake system or driving condition. That kind of claim belongs to the customer's engineering team, not to a parts supplier.

So the quote now says what is included and what is not included. If we don't have load data, a material spec, or a safety standard, the quotation says the part is for prototype or evaluation use only. That honest limitation is much better than letting someone install a shiny part and believe their brake problem is solved.

Mistakes I Still Watch For

Even with the checklist, a few errors keep trying to sneak through.

Print orientation mistakes. A printed part is not equally strong in every direction. The layer orientation has to match the main load path. If we don't consider that early, the part can fail in a way that looks like material failure but is really a design review failure.

Using worn parts as reference dimensions. If a customer brings an old worn part and says “make it exactly like this,” the measurements might already be wrong. We need the original drawing, the mating part, or a known standard.

Confusing productivity with progress. It feels good to quote quickly. It feels bad to scrap a $3,000 order. The checklist only works if we use it before the excitement of a new project takes over.

The brake pedal job taught me that lesson the hard way. We had the machines. We had the skill. What we didn't have was a clear definition of the problem. Now the checklist forces me to get one.