Before ordering, ask the supplier about mold design, achievable tolerances, resin control, machine size, mold steel, expected tool life, quality inspection, trial procedures, production capacity, maintenance, ownership, and change costs. A 30-second cycle produces 120 shots per hour before downtime; a four-cavity mold can theoretically produce 480 parts per hour. Dimensional variation also depends heavily on resin: published injection-molding guidance shows shrink-related variation can range from about 0.002 in./in. for more stable materials to 0.025 in./in. for less stable ones. The supplier should explain its numbers using your actual geometry, resin, tolerances, and annual volume rather than quoting general capability.
Start with the drawing rather than the quotation. Ask the supplier to mark every dimension that may be difficult to hold and explain why. ISO 20457:2026 covers tolerances for molded plastic parts, but a drawing still needs functional dimensions, datum references, surface requirements, and any dimensions that cannot be treated as general tolerances. A supplier that agrees to every ±0.05 mm requirement without discussing part size, resin shrinkage, gate position, cooling, or measurement method has not provided enough engineering information.
That drawing review should lead directly into a DFM discussion before steel is cut. Ask where the gate will be placed, where weld lines may form, where ejector pins will touch the part, and which walls are likely to cool unevenly. If a nominal wall changes from 2.0 mm to 4.0 mm around a boss, the thicker section holds heat longer and can increase local shrinkage or sink risk.
Ask for the proposed gate, parting line, ejector layout, draft, shrinkage allowance, and likely cosmetic defects on one marked drawing before approving tooling.
Once geometry is understood, ask how the supplier selected the resin and whether the quotation names an exact commercial grade rather than only “ABS,” “PC,” or “PA66.” Different grades within one polymer family can have different glass content, melt flow, flame rating, UV package, impact strength, and molding shrinkage. A 30% glass-filled PA66, for example, behaves differently from an unfilled PA66 because fiber orientation affects shrinkage and dimensional behavior.
Material control also needs measurable rules. Ask how incoming resin lots are identified, how opened bags are stored, whether hygroscopic materials are dried, and whether regrind is allowed. A production order for 50,000 parts should not rely on an undocumented instruction such as “dry before molding.” The purchase specification should state the resin grade, permitted colorant, permitted recycled content, and required material documentation.
From material control, move to the mold itself. Ask which steel is proposed for the cavity, core, inserts, slides, and other wear areas, plus the expected number of production cycles. A tool intended for 20,000 parts does not need the same construction as a program expecting 1,000,000 parts. If annual demand is 250,000 pieces for 4 years, the commercial discussion should address at least 1 million expected parts rather than the first purchase order alone.
The cavity count should also be supported by production math. With a 30-second cycle, one cavity gives a theoretical 120 parts per hour, two cavities give 240, and four give 480. At 20 production hours per day and 85% scheduled utilization, a four-cavity tool has far more monthly capacity than a single-cavity tool, but it also costs more and may require a larger molding machine.
| Question to ask | Number the supplier should provide |
|---|---|
| How many cavities? | 1, 2, 4, 8, or another defined count |
| Expected cycle time? | Seconds per shot |
| Planned machine? | Clamp tonnage and shot capacity |
| Mold life? | Expected cycles |
| Annual capacity? | Parts per year at stated utilization |
| Scrap assumption? | Percentage used in costing |
Machine selection should follow that capacity discussion. Ask for clamp tonnage, shot capacity, screw diameter, tie-bar spacing, and the percentage of barrel capacity used by one shot. A four-cavity mold may look economical until its projected shot weight pushes production onto a larger press with a higher hourly rate. The supplier should show the planned machine rather than stating only that “a suitable press” will be used.
After machine selection, discuss dimensional capability under production conditions rather than tool-room machining accuracy. Published molding guidance from Protolabs lists approximately ±0.003 in. machining accuracy for its injection molds while noting that molded-part shrink tolerance depends on part design and resin. The same guidance lists shrink-related variation from about 0.002 in./in. for relatively stable ABS and polycarbonate to 0.025 in./in. for TPE.
For that reason, ask the supplier to separate mold dimensions from molded-part dimensions. A steel cavity can be machined accurately while the plastic part still changes with melt temperature, holding pressure, cooling time, moisture, fiber orientation, or room-temperature conditioning. If 8 dimensions affect assembly and the drawing contains 60 total dimensions, ask for a control plan focused on those 8 rather than receiving only a general inspection statement.
Measurement should be agreed before the first trial. Ask which dimensions will be checked by CMM, optical equipment, pin gauges, micrometers, calipers, fixtures, or other methods. Also specify the sample count. Measuring 1 part from a 10,000-part batch provides very different information from measuring 5 parts at startup and another 5 parts at defined production intervals.
“We inspect every batch” is incomplete. Ask how many parts are measured, which dimensions are measured, at what frequency, with what equipment, and what happens when one result falls outside specification.
The first mold trial should therefore produce data, not only photographs and sample parts. Ask for the molding parameters used during the trial, sample quantity, dimensional report, cosmetic findings, and a list of mold corrections. If 30 trial pieces are produced, clarify whether measurements come from the first stable pieces, randomly selected pieces, or all 30. That distinction matters when judging process consistency.
Trial approval should lead into a written rule for engineering changes. Ask who pays when a dimension must be corrected because the mold was manufactured outside the approved design, and who pays when the customer changes the CAD model after tool completion. A modification involving one removable insert may take far less work than a change requiring cavity welding, re-machining, polishing, fitting, and another molding trial.
Before placing the order, compare quotation scope line by line. A tooling price of $18,000 and another of $21,000 cannot be compared properly until both suppliers state what is included.
-
DFM and mold design
-
mold steel and standard components
-
number of trial rounds included
-
sample quantity per trial
-
dimensional inspection
-
texture or polishing
-
hot-runner equipment, if used
-
spare wear parts
-
packaging
-
mold maintenance
-
engineering changes after approval
The same comparison should cover ownership. If the buyer pays 100% of the tooling charge, ask whether the contract identifies the buyer as the mold owner, whether the mold can be transferred, and whether CAD files, electrode data, drawings, or maintenance records are available. A mold with an expected life of 500,000 cycles may remain in production for years, so ownership language matters well beyond the first shipment.
Maintenance comes next because cycle count affects wear. Ask whether the supplier records total shots and schedules cleaning, lubrication, vent service, slide inspection, ejector inspection, and replacement of wear components. For a program running 300,000 parts annually, losing several production days because a small insert must be manufactured after failure can be more disruptive than paying for a spare insert during initial toolmaking.
Production records should connect maintenance with traceability. Ask whether each shipment can be tied to a molding date, machine, mold, resin lot, inspection record, and production batch. If 40,000 parts are delivered across 4 production lots and a dimensional issue appears in only one lot, batch-level records allow the investigation to focus on 10,000 parts rather than treating all 40,000 as one undifferentiated shipment.
Packaging requirements should be reviewed using the same measurable approach. Specify parts per tray or bag, pieces per carton, maximum carton weight, surface protection, allowable stacking, and labeling. If a cosmetic housing is packed 100 pieces per carton with exposed Class-A surfaces rubbing together, molding quality alone will not prevent scratches during a 5,000-mile shipment.
Lead time should also be expressed as dated stages rather than one number. A quoted “5-week mold lead time” may refer to the first mold trial, not approved production tooling. Ask for calendar dates covering DFM approval, mold design approval, steel preparation, machining, assembly, T0 or T1 trial, measurement, corrections, final approval, and production release.
When comparing a Custom plastic injection molding supplier with other candidates, use the same engineering questions and the same numeric assumptions for every quote. One supplier may quote a 28-second cycle and 2% scrap while another uses 35 seconds and 5%; those assumptions change the calculated unit cost even when resin price and machine rate are similar.
Finally, ask what happens after production has started. If monthly demand rises from 20,000 to 60,000 parts, can the existing mold meet the new requirement at 85% utilization, or will another mold be needed? If one cavity in a four-cavity tool is damaged, can production continue with three cavities? If the specified resin becomes unavailable in 2027, what approval process is required before an alternative grade can be used?
A professional supplier should be able to answer those questions with drawings, cycle estimates, sample quantities, inspection methods, capacity figures, written responsibilities, and named material grades. Ask for numbers that can be checked after production begins: seconds, millimeters, percentages, pieces per hour, sample counts, cycles, dates, and documented acceptance limits.