When I first started sourcing custom parts in 2020, injection molding felt like a legacy process—something you "graduate" from once you know what you're doing. My world was built around online ordering: upload a model, choose a material, get a price, and receive parts a few days later. If someone had asked me "why use injection molding?" I would have answered quickly: tooling costs, long lead times, huge minimum quantities.

Three years and roughly $11,000 in avoidable manufacturing spend later, I realized I had been answering a question I never actually understood.

Mistake #1: Treating a 3D printing service as the default answer

In 2022, I needed 1,200 identical PA12 parts for a new product line. I had used Sculpteo for years, mostly for prototypes and bridge pieces. The Sculpteo 3D printing service was fast, the ordering flow was frictionless, and I trusted the quality. Seeing the Sculpteo logo on the shipping label made me feel like the procurement side of my job was under control.

That comfort was the problem.

The quote was $8.20 per part. Twelve hundred parts at $8.20 worked out to roughly $9,840. At the time, that felt acceptable: a few thousand dollars for a functional product run. What I did not do was request an injection molding quote for the same geometry. I did not even think to. In my mental model, molding was only for massive consumer electronics volumes—hundreds of thousands of parts, not hundreds.

When a colleague finally asked why I had avoided tooling, the comparison was embarrassing. A short-run aluminum injection mold for that part came in around $3,900, and the per-part price at 1,200 units was about $1.15. That put the total cost at $5,280—roughly $4,500 less than the 3D printed version. The savings would only grow if we produced more than one batch.

The worst part? Nothing about the product required 3D printing. The bracket had no complex internal channels, no lattice structure, no design that additive manufacturing uniquely enabled. I had simply defaulted to the process I already used, the vendor I already trusted, and the quote that was already in front of me.

Mistake #2: Chasing a machine instead of a manufacturing process

The second costly lesson came when I tried to solve a metal fabrication problem by buying equipment instead of using a service.

I was sourcing a small metal bracket for a device enclosure. A CNC quote felt high for the quantity we needed. Around the same time, I saw a demonstration of the xTool MetalFab and started doing the math. The xTool MetalFab laser welder price, I told myself, was comparable to a few production batches. Why keep paying outside shops when the tool could sit in our workshop?

A colleague went even further. He contacted a handheld laser welding machine factory directly and got a quote that made the whole idea seem even more reasonable. The machine itself was not outrageously priced. The problem was that neither of us had accounted for the real cost of using it: operator skill, setup time, failed attempts, surface preparation, post-processing, and quality inspection for every single part.

We spent close to two weeks and several hundred dollars on experiments before admitting that a desktop tool is not a production line. The parts we made worked for prototypes, but not for repeatable, verifiable components. In the end, we still sent the bracket to a shop with the right equipment and the right operators. The xTool MetalFab might be useful for repairs or one-off custom work—that I understand now. But its price did not replace a manufacturing process. It only added a new step before I finally chose the right one.

This lesson was not really about laser welders. It was about how easy it is to confuse a machine with a production strategy.

The problem was not technology. It was my cost model.

When I looked back at both mistakes, I noticed a pattern: I was comparing the wrong numbers.

Most buyers, including me, focus on the per-unit price and the immediate tool cost. What I ignored was the full cost model across the product's life.

  • What is the total cost at 500 parts, 2,000 parts, or 10,000 parts?
  • How much does quality variation cost when parts are inconsistent?
  • How many design revisions are actually expected after production starts?
  • What are the material requirements under mechanical or thermal load?

For the 3D printed bracket, the unit price looked acceptable because I had no reference point. Once the injection molding quote was on the table, the additive quote was not just more expensive—it was more expensive without delivering any production advantage.

I also held onto an outdated belief that injection molding minimums were always enormous. That legacy myth comes from an era when tooling was steel, mold design was entirely manual, and prototyping a mold meant waiting months. The manufacturing world has changed. Hard tooling is still used, but aluminum tooling, conformal cooling, simulation, and online DFM tools have shortened the process significantly. What was best practice in 2020—or in 1990—is not necessarily the right assumption in 2025.

What actually changed my approach

Today, I still use Sculpteo's 3D printing service for prototyping and low-volume production. The difference is that I no longer treat it as the default. The Sculpteo logo in my inbox now reminds me that the platform can do more than one process. It also serves as a place where I can compare the economics of additive, CNC, and injection molding on the same part data.

My own practical rule is now simple: if a part is expected to live beyond 500–1,000 units over its lifetime, I request an injection molding quote before I commit to any other process. Not because injection molding is universally better—it is not. But because the decision should be based on total cost, not on which ordering interface is most familiar.

There are still parts that belong on a 3D printer: highly complex geometries, impossible-to-machine internal features, very small quantities, and design iterations that are still changing. And there are still parts that belong on a CNC machine. But for a plastic component that needs consistent material properties, repeatable tolerances, and reasonable unit cost at volume, injection molding answers a question no amount of machine shopping can solve.

The expensive lesson, summarized: the goal is not to avoid old processes or collect new tools. The goal is to match the process to the volume, the material, and the quality requirement. If I had asked that question in 2022 instead of trusting the easiest quote, I would have saved thousands of dollars and a lot of unnecessary frustration.