Maybe Additive Is Too Expensive

The verdict on additive manufacturing is often that it is impressive but too expensive, and plenty of cost-per-part comparisons confirm it. Using a comparison standard written for conventional processes, the math is sound. That makes the verdict hard to argue with, and easy to accept. Even when it is wrong.

Begin with the question being asked: what does this part cost? It is a fair question, and a necessary one. The trouble is that the standard comparison answers it incompletely. It counts what happens at the production step and leaves out real costs that belong to that part but fall outside what the comparison includes. The resulting cost estimate is not so much wrong as partial – and once the rest of the cost is counted, additive often looks very different.

The comparison was built for a narrower job. It was designed to weigh conventional processes against each other, where everything upstream and downstream is assumed to be the same and only the production step differs – the impacts diverging and converging along a linear production path. So the unit of measure defaults to one part from each method, and the boundary is drawn tight around production. Even a lifecycle assessment that reaches into the use phase – fuel burned, energy consumed – usually captures only a slice of what happens after the part leaves the line. That narrow frame works when the two options really are interchangeable ways to make the same part. Additive is not that. The costs it changes ripple outward – at the part itself, in the system the part sits in, and across the enterprise around it – and a part-to-part comparison drawn that tightly captures almost none of it.

Every transformative technology has initially been misunderstood in its potential and mismeasured in its impact. Desktop computing comparisons initially measured a PC against IBM’s best typewriter. The math worked – it showed a computer was an expensive way to produce a document. What the comparison missed was that the computer was not a better typewriter or simply another business tool but a different kind of business system, and soon to enable a completely different way of organizing work and doing business. Artificial intelligence is being argued in the same terms right now – simply doing current work better or enabling something else entirely. The comparisons can provide the right answer to the wrong question. Maybe additive needs better questions.

Four gaps show where a careful, well-intentioned cost comparison stops short of the full cost of the part.

The part is often better. An additively produced part is frequently lighter, stronger, or more durable, and the effects run downstream: better performance, less maintenance, fewer replacements. That value is real and measurable, and most of it never enters the part-cost comparison, which assumes both methods converge on the same resulting part. For additive, they often do not.

Durability especially compounds in a way the standard comparison is not built to capture, because of the unit it measures. The functional unit – the basis on which the two options are compared – is the least examined choice in most analyses, and it usually defaults to one part against one part, even when the service life of the system would be far more informative. One part is simply easier to measure, so it rarely gets questioned.

An assessment of an aircraft component showed this even with careful analysis. The study compared the part machined conventionally against the same part produced additively, measuring the material, energy, cost, and emissions of each, and then the in-service fuel impact of the lighter additive part against the heavier conventional one. It found clear fuel and emissions gains from the lighter part, and negligible difference in the production phase itself, part for part. What it did not account for was durability: subsequent analysis suggested the additive part was significantly more durable, with a longer service life. The study used one part as its functional unit. Had it used the life of the aircraft instead – six conventional parts against four additive ones – the production comparison would have come out very differently.

A second gap is where the analysis stops. The boundary is drawn at the point where the two production paths are assumed to converge, or simply where the data runs out. Draw it around production alone and everything additive changes downstream falls outside it: inventory never produced, warehoused, and written off; a design left open instead of locked early, with the improvements that follow; more durable parts needing fewer replacements. Real costs, real savings, real strategic effects – left out not because they are small, but because of where the line was drawn.

A third gap is how cost gets attributed when a single part is really tied to a minimum production run. A heavy-equipment maker supports its legacy machines as a matter of brand, supplying spare parts for old equipment at a loss. Getting one such part from a conventional supplier may require a minimum order of 100. One ships to the customer; the other 99 are scrapped, or warehoused against demand that may never come. The additive version has a minimum of one. So what does the conventional part cost in the comparison – one, or 100? In most analyses it carries the cost of one, and the write-off for the other 99 lands on a different ledger or a later quarter. But those 99 were a cost the single needed part caused, and even warehoused they tie up capital and carry risk. The accounting is standard. The comparison is still wrong.

A fourth gap has nothing to do with manufacturing at all – it is how tooling gets budgeted. Conventional tooling is often expensive enough to be capitalized: depreciated over years, approved on a separate track, charged to a separate account, and absent from the departmental budget and the part-cost comparison. Additive’s equivalent cost is often small enough to fall below the capital threshold, so it is treated as an operating expense and shows up directly in the comparison. The more cost-effective option can look worse to the person managing a departmental budget – not because it costs more, but because of where its cost appears.

And that is not a quirk of one company’s books. Capital depreciation exists precisely because conventional manufacturing’s tooling is long-lived and expensive, and its cost has to be spread across years. The accounting conventions, the capital thresholds, the tax code – all were shaped around that conventional cost structure. It is a cost structure additive does not share.

The whole apparatus – the standard analysis, the cost allocation, the boundaries – was built around conventional manufacturing’s economics. Additive’s business impacts are being measured with instruments calibrated to making parts. In every case the accounting is correct on its own terms. But correct accounting can still leave out real cost.

After one such evaluation was presented, an analyst was asked how the downstream impacts had been accounted for. A little indignantly, the answer came back: “That’s really hard to calculate!” True enough. But hard to calculate is not the same as unimportant. Left out of a comparison presented as complete, those impacts still shape the decision – they just do it invisibly, steering it toward the wrong answer.

Follow the costs all the way out and the largest of them are not at the part level at all. This is where choosing additive stops being a part-cost question and becomes a business decision – because the same part-level view misses what additive enables beyond production entirely: bringing products to market on timelines conventional production cannot meet; optimizing each part for its purpose instead of standardizing components to bundle volume; keeping a design open until demand is clear; responding to shifts in the market that conventional constraints would have closed off.

Additive manufacturing is at an inflection point, and a moment of real opportunity. Infrastructure investment, reshoring, policy frameworks, procurement specifications – all are being written now, most of them on a part-level view of cost that misses where much of additive’s value lies. Whether additive is treated as a slightly different way to make the same part, or as something that changes how a business operates, depends on measuring it against the right question. The decisions being made now will shape the next decade of manufacturing. This is the moment to make them with the full picture.