3D Printing vs. CNC vs. Casting: How a Manufacturing Buyer Chooses the Right Process

The Three Manufacturing Scenarios I Sort Every Request Into

I am the office administrator and buyer for a 40-person product development firm. I manage our prototyping and low-volume manufacturing purchases—roughly $200,000 a year, maybe $180,000 depending on what finance counts as tooling. I am not an engineer, so I do not pretend to choose materials. What I do is ask which manufacturing process fits the part, then get quotes from suppliers. After five years of this, I have learned that there is no universal default process. There are only scenarios.

Most requests fall into one of three buckets:

• Utility parts. The part is for our own workshop, desk, or office. Nobody outside the company will ever see it.
• Functional test parts. The part must survive real use and produce reliable data before the design gets approved.
• Customer-facing parts. The part ships with a product or is shown to a client. Its finish directly affects how the customer sees our company.

Here is how I handle each scenario.

Scenario A: Internal Utility Part — Download the File and Print It

A few months ago, one of our engineers walked in with a Milwaukee M12 tool holder 3D print file he had found on a model-sharing site. It was a wall mount for his M12 batteries, charger, and drill. He asked whether we should make it through a proper manufacturing vendor or just buy an injection-molded holder online.

I asked my usual first question: What happens if it breaks? He said nothing. It would just get reprinted. The holder was for organizing a workbench; nobody would get hurt and no customer would ever see it. So we went with 3D printing. We printed that Milwaukee M12 tool holder 3D print file in PETG on an in-house printer for about $6 in material. If we did not have a printer, I would have sent the same file to a local print service and probably spent less than $25. It did not need a full manufacturing review because it did not need one.

This is the scenario where downloadable 3D print files shine. The M12 holder was our version of a low-risk utility part. Trying to over-engineer procurement for these requests is a waste of time and money.

One caveat: this advice depends on the word internal. Even a small tool holder becomes a different problem if it holds something heavy above a person or sits near heat. In our workshop, the conditions were mild. Your situation may be different.

Scenario B: Functional Part — Casting Is Not Additive Manufacturing, and 3D Printing Is Not Always Cheapest

This is the scenario that changed my mind as a buyer.

At a design review in early 2025, our project manager asked, in all seriousness, is casting additive manufacturing? I had the same question when I first started, because vendors throw around terms like digital manufacturing and rapid prototyping. The short answer is no. Additive manufacturing builds an object layer by layer from a 3D CAD file. Casting is a separate process where molten material is poured into a mold cavity. Both start with a CAD file, but the process economics are not interchangeable.

The prototyping services via 3D printing market has grown to the point where ordering a printed part is the easy part. That convenience creates a habit: defaulting to additive whenever a CAD file exists. But additive is not always the cheapest or fastest route to a functional metal part. Often, CNC machining beats it.

Here is a real example. In late 2024, our engineers needed 30 aluminum brackets for a customer field trial. The bracket was not geometrically complex: flat faces, a few holes, and a mounting slot. We quoted two processes. The CNC quote came back around $600 for 30 parts with a six-business-day lead time. The metal 3D printing quote for the same CAD file was closer to $1,800, maybe $1,900—I would have to dig up the exact quote—because support removal and inspection are expensive. I did not choose 3D printing. I chose CNC machining.

This feels backwards to many people because metal 3D printing is often presented as the future of manufacturing. In reality, for a simple bracket machined from a solid billet, additive manufacturing has few advantages. CNC is faster, cheaper, and produces material properties that match the datasheet. When engineers are evaluating a part before production, that consistency matters more than process novelty.

When I need to compare options, I use the Fictiv website. The Fictiv website accepts a CAD upload and returns quotes for CNC machining, injection molding, and 3D printing in one place. I can see the price difference in 10 minutes instead of chasing suppliers by email. Fictiv also tracks the order digitally, which helps when finance asks where a purchase order is. In fact, the Fictiv manufacturing platform is where our company starts almost every custom part request now.

Prototyping services via 3D printing are still part of our workflow. For a non-structural housing, a complex duct, or an ergonomic model, additive is excellent. My point is narrower: the easiest request is not always the best process. Quote both.

Scenario C: Customer-Facing Part — Finish Quality Is Brand Quality

This is where I learned the most painful lesson of my purchasing career.

In 2024, we produced 20 custom enclosures for a client pilot. I approved an online 3D printing order because the quote was about $300 lower than the production-grade option. I was proud, thinking we had saved money without affecting function.

The enclosures arrived with visible layer lines, a rough texture, and inconsistent color. Functionally they were fine. Visually they told the client that we ship unfinished work. The client engineer did not need to say much; the look on his face said everything. We then paid another $400 for sanding, priming, and re-finishing and still apologized for being late. The $300 saving turned into a $400 loss plus damaged client trust. I now use that project as my personal reminder that parts with customer visibility deserve a different standard.

When a part is customer-facing, quality is not an upsell. The first physical contact a customer has with your product is the finish. A part with rough edges signals weak engineering. A part with clean surfaces signals attention to detail. In my opinion, spending more on a production-appropriate process is justified when your brand is on the line.

For low-to-mid volumes, this usually means CNC machined parts with anodizing or painting, or a well-post-processed 3D print from a service provider you trust. Above a few hundred units, injection molding is often the right call because the tooling investment pays back through speed, repeatability, and surface finish. For high-volume metal parts, casting can also be a legitimate production process, but it belongs to a different discussion and typically requires a foundry rather than an instant online quote.

How to Tell Which Scenario You Are In

If you are not sure which bucket your part belongs to, ask four questions before requesting a quote.

1. What happens if the part breaks? If the answer is inconvenience, it is Scenario A. If the answer is a failed test or a delayed project, Scenario B applies. If a customer will notice, it is Scenario C.
2. How many parts do we need? One copy of an internal tool holder is a print-and-forget job. Dozens of functional parts can be machined or printed depending on geometry. Hundreds of customer-visible parts normally point toward injection molding or casting.
3. Who will judge the quality? Your own team judges Scenario A. An engineer measuring tolerances judges Scenario B. A customer making a purchasing decision judges Scenario C. That changes how much finish matters.
4. What is the deadline? A functional metal bracket can often be machined faster than a metal 3D print can go through build, support removal, and post-processing. Do not assume additive is quicker just because it sounds modern.

In our company, the workflow now looks like this. An engineer gives me a CAD model and an intended use. I upload it to the Fictiv website to get baseline quotes for CNC machining, 3D printing, and injection molding. Then I compare those numbers with the answers to the four questions above. Sometimes we 3D print it. Sometimes we machine it. Sometimes we build a mold.

I can only speak from my experience in product development. If your company makes safety-critical components or regulated medical devices, your requirements will be stricter, and this general framework is not enough. But for the everyday question of how to turn a CAD file into a physical part, the scenario approach works: match the process to the purpose, not to the hype.