Additive manufacturing solves real problems. It also gets reached for when it shouldn't be, and the cost of that mistake rarely shows up on day one. Here's how to recognize the wrong fit before it recognizes you.
A part that started as a fifty-unit pilot run is now on order number four, and the quantity just jumped to five thousand. Nobody in the room asked whether the manufacturing process should change along with it. The part has printed successfully every time so far, the supplier relationship already exists, and switching feels like unnecessary risk on something that isn't broken. Six months later, the margin on that part quietly turns negative, and the postmortem traces back to a decision that was never actually made. It just kept not being reconsidered.
This post is the uncomfortable one in the series. Fourteen posts in, it would be easy to keep building the case for additive manufacturing as a general-purpose answer. It isn't one, and pretending otherwise is how good engineering teams end up with a part that quietly blew the budget or missed the shipping window. The goal here isn't to talk anyone out of the technology. It's to build the same kind of instinct for recognizing a bad fit that experienced buyers already have for a good one.
Every manufacturing process has a zone where it wins and a zone where it loses, and additive manufacturing is no exception, no matter how flexible it feels on the shop floor. The zone it wins in is real: low to moderate volumes, geometry that would be expensive or impossible to produce any other way, and situations where the cost of tooling would dwarf the cost of the parts themselves. Outside that zone, the same properties that make additive manufacturing remarkable, slow, part-by-part production and material flexibility, turn into liabilities. The mistake isn't choosing additive manufacturing. It's choosing it out of habit or convenience rather than out of fit, and not checking that fit again once the volume, tolerance, or certification requirements change.
Teams that get early wins with additive manufacturing on prototypes often keep using it well past the point where a different process would be faster and cheaper, simply because the workflow is already set up and familiar.
Instead of a checklist, it helps to hold a single mental picture, one worth calling the Fit Zone: two axes, one for production volume and one for how much the part's geometry or customization needs actually demand additive manufacturing's freedom.
Plot any part on that grid. Low volume paired with genuinely complex or customized geometry sits in the sweet spot: this is where additive manufacturing wins outright, because tooling for any alternative process would cost more than the entire production run. High volume paired with simple geometry sits in the opposite corner, and that corner belongs to injection molding, stamping, or die casting almost every time. The two remaining quadrants are where the real judgment calls live. High volume with high complexity often points toward a hybrid approach, printing masters or inserts that feed a faster downstream process. Low volume with simple geometry is a genuine toss-up, and computer numerical control (CNC) machining frequently wins there on lead time, surface finish, and material properties, even at low quantities.
A quick example makes the grid concrete. A custom bracket for a one-off robotics prototype, built in small numbers with a shape driven entirely by the space it has to fit into, sits comfortably in the sweet spot. A simple enclosure for a consumer device shipping in the hundreds of thousands sits just as comfortably in the opposite corner, and no amount of design cleverness moves it out of that corner once the volume is that high. The interesting cases, and the ones worth actually pausing on, are the parts that sit near the boundary lines rather than deep inside a quadrant. Those are the parts where the answer genuinely depends on specifics like tolerance stack-up, expected order growth, and whether the part's geometry could be simplified enough to make a tooled process viable.
The diagram below lays out these four zones with the reasoning behind each one, and it's worth sitting with before applying it to a specific part. The framework in Design Trade-offs: Strength, Speed, and Cost covers the tradeoffs within additive manufacturing itself; this one is about the decision to use it at all.
In practice, three signals show up over and over when a part has drifted out of the fit zone, and none of them are dramatic on their own.
The first is a volume crossover that nobody re-checked, like the pilot-to-production jump described at the start of this post. Additive manufacturing's cost per part barely improves with volume the way injection molding's does, since there's no tooling cost to amortize away, only build time that stays roughly constant per unit. A molded part, by contrast, starts expensive because of the tooling investment and then gets cheaper with every additional unit, since that tooling cost gets spread across a larger run. The two cost curves cross somewhere, usually in the low thousands of units depending on part size and geometry, and past that crossover point the molded or machined part comes out ahead even after its tooling is fully paid for. Nobody has to make a dramatic mistake for a part to end up on the wrong side of that line. The order just has to grow, and nobody has to remember to check.
The second signal is a tolerance or surface finish requirement that additive manufacturing structurally can't hit without secondary processing. Sealing surfaces, bearing fits, and optical-quality finishes often need machining, polishing, or coating after printing regardless of which process built the part, and once those secondary steps are added, some of the labor and lead-time advantage that made additive manufacturing attractive in the first place disappears. If a part needs machining anyway, it's worth asking honestly whether the print itself is adding value or just adding a step.
Some tolerance and finish requirements are worth flagging as an immediate no, not just a factor to weigh. A sealing face that needs to hold a fraction of a degree of flatness across a wide surface, a bore that needs a tight sliding fit straight off the machine, or a cosmetic surface that has to be free of visible layer lines without any secondary finishing are all cases where additive manufacturing's as-printed capability falls short on its own terms, independent of cost or volume. When a spec sheet includes requirements like these, the honest move is to treat additive manufacturing as disqualified for that surface from the outset, rather than hoping a good printer or a skilled operator will close the gap.
The third is a material qualification burden that outweighs the manufacturing benefit. Aerospace and medical parts frequently require certified material batches, documented process parameters, and repeatable mechanical properties across every single part. Additive manufacturing can meet those bars, but doing so is expensive and slow to set up, and for a part that doesn't actually need the design freedom additive manufacturing offers, a qualified traditional process with an existing certification history is often the faster route to a shippable product.
If a part can be 3D printed, it should be." Printability says nothing about whether additive manufacturing is the right process for that part at that volume, with that tolerance, and under that certification requirement. Those are four separate questions, and each one can independently rule the process out.
Traditional processes like injection molding remain the default for exactly the volume and geometry combination additive manufacturing struggles with. It's worth looking at what that alternative actually involves before assuming additive manufacturing is automatically the more modern or more efficient choice.
Here's the part that doesn't get said enough: additive manufacturing rarely fails loudly. Nobody's printer explodes when it's the wrong choice. Instead, the part ships, it works, and the mismatch shows up three months later as a margin that never materialized, a production line that can't scale past a certain order size, or an audit that flags missing material certification. The technology is forgiving enough in the moment that the wrong decision doesn't announce itself until it's expensive to reverse.
The teams that get this right treat the process decision as something to revisit, not something to set once. A part's volume, tolerance requirements, and certification needs can all shift after the initial manufacturing choice was made, and the discipline is in checking the fit again rather than defending the original decision because switching processes feels like admitting a mistake. It isn't a mistake to have started with additive manufacturing for a prototype. It's a mistake to still be using it for the same reason a year and ten thousand units later.
This is also where cost modeling earns its keep. The cost iceberg framework makes the point that a lot of additive manufacturing's real cost is hidden below the obvious material and machine-time line items. That same iceberg is exactly what makes the wrong-process mismatch so easy to miss in the moment and so expensive once it surfaces. A part that looks fine on a per-unit cost spreadsheet can still be the wrong choice if the spreadsheet only counts material and machine time and skips post-processing labor, quality inspection, or the engineering hours spent re-qualifying a design every time the printer or material batch changes.
None of this means additive manufacturing needs a business case bulletproof enough to survive an audit before anyone's allowed to use it. Plenty of good decisions get made with incomplete information and a reasonable read on where a part sits in the fit zone. The point is narrower: build in a moment, ideally tied to a volume milestone or a design freeze, where someone actually asks the question again instead of assuming the original answer still holds.
Think about custom tailoring versus buying a shirt off the rack. A tailor is worth every dollar for someone with a build standard sizing doesn't fit, or for a single event where the fit has to be exact. Nobody would seriously suggest getting a hundred identical shirts custom-tailored one at a time when a factory running standard sizes could produce them faster, cheaper, and with more consistent quality. The tailor isn't a worse crafts person than the factory. The two are simply solving different problems, and the mistake is applying the tailor's method to a job the factory was built for. Additive manufacturing is the tailor. It's excellent when the part genuinely needs that individual attention, and it's the wrong tool the moment the job turns into a hundred identical shirts.
Additive manufacturing wins in a specific zone: low to moderate volume paired with geometry or customization that would be expensive to tool for any other way. Outside that zone, other processes usually win.
Watch for volume creep. A part's production quantity can grow well past the point where additive manufacturing still makes economic sense, and nobody may have flagged the crossover.
Tolerance, surface finish, and material certification requirements can each independently rule additive manufacturing out, regardless of how well the part prints.
The wrong process choice rarely fails loudly. It shows up later as missed margin, a scaling wall, or a certification gap, which is exactly why the decision deserves revisiting rather than defending.
Printability is not the same question as suitability. A part being 3D-printable says nothing about whether it should be.
Next week closes out the season with a look at where all of this leaves the reader: not a verdict on additive manufacturing, but a framework for deciding, part by part, when it earns its place on the floor and when it doesn't.
