You have a prototype that works. Turning it into thousands of consistent, sellable parts is its own stage โ with its own steps and its own expensive mistakes. Here is the real path, and how to avoid the traps.
The short answer: move in stages, not one leap. Lock the design with a manufacturability review; choose the right tooling (prototype or bridge tooling for modest volumes, a production tool once design and demand are proven); run a first-article inspection to confirm the first real parts match the drawing; do a small pilot run to shake out problems; then go to full production with an agreed quality plan. The costly mistake almost everyone is tempted by is jumping straight from a prototype to a production tool while the design is still changing โ do that and the tool needs rework that costs far more than the staging would have.
A prototype and a production part are made differently, and that difference is the whole challenge. A prototype is often 3D printed or machined, in ones and twos, sometimes in a stand-in material. A production part is made by the final process โ usually injection molding for plastics โ in the final material, thousands of times, and it has to come out the same every time. A shape that printed fine can behave differently when molded: it may shrink, warp, or reveal a feature that is hard to tool. So the move to production is not just "make more of the prototype" โ it is re-making the part in the real process and proving it stays right at volume.
Production tooling assumes a finished design, so settle it first. This is the point for a serious design-for-manufacturability review: uniform wall thickness, draft so parts release, no unnecessary undercuts, features that the molding process can actually hold. Every change you make after the production tool is cut is slow and expensive, so the goal is to walk into tooling with a design you are confident in. If the design is still moving, that is a sign to stay on prototype or bridge tooling a little longer, not to commit the production tool.
A 3D print proves shape and fit, but a molded part behaves differently โ and the DFM review exists to close that gap. The things that change (and that a maker will raise before cutting steel):
None of this means a print is useless โ it's the fastest way to prove shape. It means the step between a good print and a good mold is real work, which is exactly what a production-minded review handles up front.
Match the tool to how proven you are. Prototype or bridge tooling โ often aluminum or softer steel โ makes real production-material parts in modest numbers at lower cost, ideal when you want to sell, test, or launch before committing. Production tooling โ hardened steel, sometimes multi-cavity โ costs more and takes longer but runs high volumes consistently over a long life. Many products use both in sequence: bridge to launch and learn, production once the design is locked and volume is real. See what is bridge tooling and prototype vs production tooling.
When the new tool makes its first parts, do not run thousands โ inspect the first articles against your drawing. A first-article inspection measures the key dimensions and features and confirms the parts match the spec before full production. This is your cheapest opportunity to catch a tooling or process problem: a dimension out of tolerance, a cosmetic flaw, a feature that did not fill. Take first-article approval seriously and in writing โ it is the gate that decides whether production parts will be right.
First shots are iterated, not final. The usual rhythm is T1 (first samples off the new tool) → measure and review → corrective steel adjustments → T2 samples → approval, sometimes a T3. This is normal, not a failure. It's also why tools are designed steel-safe: a toolmaker can remove metal easily to open a dimension or enlarge a feature, but adding metal back (to tighten one) usually means welding or a new insert โ expensive. So the tool is cut deliberately on the side that leaves room to adjust, and the sampling rounds dial it in.
A pilot run โ a small production batch โ shakes out the problems that only appear at real speed and quantity: cycle-time issues, occasional defects, assembly snags, packaging that does not quite work. It is far cheaper to find these on a few hundred parts than on a full run of thousands. Use the pilot to confirm the process is stable and to finalize your packaging and assembly before you scale.
Going to volume means agreeing, up front, how quality will be checked โ not hoping for the best. That usually means an inspection standard: a sampling plan that defines how many units are inspected and how many defects are acceptable before a lot is rejected. The common standard for this is AQL sampling (ANSI/ASQ Z1.4), and you can work out a plan with our free AQL calculator. Agreeing the standard before production protects both sides and keeps defects from reaching your customers. Reorders then run from the same proven tool and process โ see reorders and repeat production.
Stage-gate at a glance (what each stage produces and who signs off):
| Stage | Deliverable | Pass criterion |
|---|---|---|
| Lock design | Final CAD + drawings, DFM review | Design frozen; no open changes |
| Tooling | Mold cut; first shots (T0/T1 samples) | Parts fill and release cleanly |
| First article | FAI report: ballooned drawing, key dimensions measured, material cert | Dimensions in tolerance; you approve in writing |
| Pilot | Small production batch | Process stable; defects understood |
| Production | Full runs + inspection to the agreed standard | Lots pass the AQL plan |
Worked AQL example: at General Inspection Level II, a lot of ~5,000 parts maps to a sample size code letter that sets how many you inspect; at AQL 2.5 the plan gives an accept/reject pair (accept the lot at N or fewer defects, reject at N+1). You also classify defects as critical / major / minor, often with a stricter AQL (or zero) for critical. Build your exact plan with the free AQL calculator โ enter lot size, level, and AQL and it returns the sample size and accept/reject numbers.
What to hand your manufacturer: final CAD (STEP) + drawings with tolerances, the material/grade spec, your target annual volume and budget, any certifications the market needs, and who owns the tooling โ get tool ownership in writing (see tooling ownership).
Worked AQL example (ANSI/ASQ Z1.4, normal single-sampling): for a lot of 5,000 parts at General Inspection Level II, the standard gives sample-size code letter L โ inspect a random 200 parts. At an AQL of 2.5, you accept the lot on 10 or fewer defective parts and reject on 11 or more. For critical defects (safety), suppliers use a much tighter AQL โ often 0 (accept only on zero). These are the published Z1.4 table values; confirm the exact plan with your inspector.
JCH manages the whole transition โ DFM, tooling, first articles, pilot, and a quality plan โ with one point of contact. Send Jack your prototype or drawings. NDA first, no job too small.
Book a 30-min callMake a prototype firstLock the design first, then move through a staged path: a design-for-manufacturing review, the right tooling (prototype or bridge tooling for modest volumes, production tooling once the design and demand are proven), a first-article inspection to confirm the first real parts match the drawing, a small pilot run to shake out problems, then full production with a quality plan. Rushing straight from a prototype to a production tool on a design that is still changing is the most common and most expensive mistake.
A prototype is made to learn โ often by 3D printing or machining, in small numbers, and it may not use the final process or material. A production part is made by the final process (such as injection molding) in the final material at volume, holding consistent dimensions part after part. Because the process and material differ, a part can look right as a prototype and behave differently in production, which is why a pilot run and first-article inspection matter.
Not necessarily. If your volume is modest or the design is not fully proven, prototype or bridge tooling makes real production-material parts at lower cost and risk while you confirm demand and finalize the design. You step up to a hardened production tool once the design is locked and the volume justifies the investment. This staged approach keeps your spending in line with how proven the product is.
A first-article inspection (FAI) is a detailed check of the first parts off a new tool against your drawing and specifications, measuring the key dimensions and features before full production runs. It catches tooling or process problems while they are still cheap to fix. Approving first articles is a normal, important step you should expect and take seriously, because it is your checkpoint that the production parts will be right.
The common problems: tooling a design that is still changing, so the tool needs expensive rework; discovering that a prototype material behaved differently than the production material; skipping first-article inspection and finding defects after a full run; and having no quality plan, so defects reach customers. Nearly all of these are avoided by locking the design, running a pilot, inspecting first articles, and agreeing an inspection standard up front.
Usually no for volume โ printing is slow and costly per part and limited in material. You move to molding or the right production process, which is why the staged path exists. See molding vs 3D printing.
Agree an inspection standard (such as AQL sampling) up front and inspect to it. The free AQL calculator builds the plan.
Minor changes can sometimes be made to a tool; major ones may need rework or a new tool. This is exactly why locking the design before production tooling saves the most money.