Prototyping a Product: A Step-by-Step Guide

A prototype is not a small production run. It is a question you are asking of the physical world, and the useful ones ask one question at a time. What follows is the route most products take from an idea to a batch you can sell. The reason to lay it out in stages is that each stage has a different job, a different material and a different definition of good enough — and most wasted money comes from doing a later stage's work at an earlier stage.

Stage one: decide what the print has to prove

Before any CAD, write one sentence: what does this print need to tell me? It is almost always one of four things.

• Does it fit? Against an existing panel, tube, device, shelf or hand.
• Does it work? Does the mechanism move, the clip hold, the lid stay shut.
• Does it look right? Proportion, colour, weight, feel.
• Does it convince someone? A model for a buyer, a design review, a crowdfunding video.

These want genuinely different prints. A fit check can be a fragment — just the mounting face and the holes, printed in twenty minutes. An appearance model can be hollow and mechanically useless but needs sanding and paint. Deciding the question stops you paying for a full functional part when a coupon would have answered it. The same logic drives architectural models, where a massing block settles what no render can.

Stage two: get it into CAD

Printing needs a solid model — a watertight, manifold body, not a surface sketch. If you model your own, design in adjustment from the start: slots instead of holes, chamfered lead-ins, screws rather than fits that must be perfect. If you have no CAD and a good idea, that gap is what 3D design help is for; a sketch with dimensions and a description of what the part does is enough to start.

Stage three: the first print is a reality check

Expect the first one to be wrong. That is what it is for. Print it in PLA — it is cheap, fast, dimensionally well-behaved and forgiving to print, and its brittleness does not matter yet because you are not testing strength, you are testing whether you drew the right thing.

Then be systematic about what you learn. Put calipers on it. Check the things that must fit against the thing they must fit. Nothing comes out at nominal size — holes shrink, posts grow — so read tolerances and fit before you conclude the design is wrong when it is actually the clearance.

Stage four: iterate, one change at a time

The discipline that makes prototyping cheap: change one variable per round, and print the variants together rather than one after another. Two rounds that each teach you something beat five rounds where you changed four things and cannot tell which one helped. How many prototype iterations covers the method and the realistic count — for a simple bracket, three or four rounds; for anything with a mechanism, a seal or a human hand on it, more.

When to stop printing in PLA

PLA is a modelling material. The moment your question changes from "is this the right shape?" to "will this survive?", change material.

• PETG — the sensible default for a functional prototype. Tougher, bends before it breaks, tolerates a warm car or a window sill.
• ASA or ABS — outdoors, UV, sustained heat.
• Nylon or carbon-fibre nylon — living hinges, gears, real engineering loads.

Test in the material you intend to ship in, or the test is fiction. A PLA part that passes a drop test tells you nothing about the PETG version, and vice versa. PLA vs PETG vs ABS sets out where each one fits.

Stage five: the pre-production batch

Once the design stops changing, print a batch — commonly tens rather than hundreds — and put it in front of real users, or sell it. This is where the problems no prototype finds turn up: assembly that takes too long, packaging, a tolerance that works once and not thirty times. From prototype to production covers this bridge and the point where tooling wins instead.

Where this is the wrong process

FDM is not the right answer everywhere in this arc, and it is worth knowing where it stops:

• Appearance models needing a flawless moulded finish. FDM leaves layer lines. They sand and fill out, but that is hand labour, and a mirror finish is real work.
• Fine detail below roughly a millimetre. Nozzle diameter sets the floor. Tiny text, hair-fine detail and thin lattices are not FDM's fight — and we print FDM only, so we cannot offer you a resin alternative.
• Optically clear parts. Printed "transparent" filament is translucent at best.
• Testing a moulded part's exact material behaviour. A printed PETG part is not a moulded PETG part: it is anisotropic and has layer lines as potential failure planes. See how strong are 3D printed parts.
• A settled design at volume. If nothing has changed in three rounds and you need thousands of units, stop prototyping and get a tooling quote.

If your part sits in one of those, we would rather say so at the quote stage than take the order.

Get an estimate · see the product prototyping service · how pricing works.