Two different questions, two different tools

A prototype exists to answer a question. In a product program there are two very different questions: does it work? and will people want it? Trying to answer both with a single unit is one of the most expensive early-stage mistakes there is. You end up fighting engineering problems and appearance problems at the same time, where every change on one side breaks the other, and progress slows to a crawl while the cost per iteration climbs.

Professional practice splits them into two parallel tracks: a works-like prototype that proves the technology, and a looks-like prototype - also called an appearance model - that proves the form. The short answer to which one you need: build the works-like first if your biggest open risk is technical, and the looks-like first if your biggest open risk is whether anyone wants it or whether you can raise money to find out.

Works-like: proving the technology

Its job is to demonstrate that the mechanism lifts the load, the sensor detects reliably, the battery lasts a day, the algorithm converges. It is allowed to be ugly - a development board with jumper wires, a hand-cut aluminum chassis, a coarse printed housing held together with tape.

Every dollar spent making it attractive is wasted, because it is going to change ten more times. What it should get instead is instrumentation: test points, logging, and the ability to swap subassemblies quickly. The distinction between this and an even earlier experiment is drawn in proof of concept vs prototype. On a later revision, this is also the unit that carries the first durability work described in reliability testing for a new product.

Looks-like: proving the appeal

Its job is to show the finished product: form, color, finish, weight, and hand feel - with nothing functional inside. It is typically built by high-resolution resin printing or machining, then filled, sanded, primed, painted, and assembled to a cosmetic standard, sometimes weighted with internal mass so it feels like the real thing. Details of that finishing work are covered in painting and finishing prototypes, and the same technique produces the non-functional mockups used for early form reviews.

It has three main uses: user testing of ergonomics and appeal, early marketing photography, and meetings with investors, retailers, and licensees - where a unit that looks real changes the entire conversation, as discussed in how to demo a prototype to investors. For some of those uses, a photorealistic rendering does the job at a fraction of the cost; the comparison is in product rendering cost.

Why separating them saves money

  • Each track moves at its own pace. Engineering can churn through ten cheap revisions while industrial design polishes two expensive ones.
  • Each uses the cheapest technology for its own purpose. Fast, rough printing for function; fine resin printing and hand finishing for appearance.
  • Failures stay cheap. When a mechanism breaks in testing, a week of paint work does not break with it.
  • Decisions get clean answers. A user reacting to an appearance model is reacting to the form, not to the exposed wiring.

The convergence point

The two tracks merge when their answers stabilize - the technology is proven and the form has been validated with users. Only then do you build the first integrated unit, often called a beta prototype or pre-production unit: it both works and looks right, and it is built as close as possible to the final manufacturing processes.

That unit is expensive, so build it once rather than five times. It is the basis for certification testing and for the pilot production run, and it sits at the boundary described in EVT, DVT, and PVT. The most common scheduling error in early-stage hardware is building the integrated unit too soon and then paying for repeated revisions of the most expensive artifact in the program.

Cost orders of magnitude

Absolute numbers depend on complexity, but the ratios are stable:

  • First works-like prototype: a few hundred to a few thousand dollars when it is assembled from off-the-shelf modules and in-house prints. Tens of thousands once professional engineering time, custom boards, and custom mechanisms are involved.
  • Quality appearance model: low thousands to tens of thousands of dollars depending on size, part count, and finish standard. Most of that is skilled hand labor - sanding, painting, and delicate assembly are craft hours, and craft hours are expensive.
  • First integrated unit: more than the sum of the two tracks, because it needs engineering, cosmetic finishing, and the reconciliation work between them.

Detailed breakdowns for both tracks are in cost to make a prototype. The budgeting rule that follows: many cheap iterations within each separate track, very few expensive integrated units at the end.

Which to build first, concretely

Ask which unknown would kill the project. If you do not yet know whether the core mechanism is feasible, spending on appearance is premature - and possibly wasted entirely. If the technology is straightforward and the real risk is market demand or fundraising, the appearance model is the higher-return purchase, and it is also what unlocks the pre-order and demand tests that de-risk everything downstream.

Framed in risk terms: the works-like track removes technical risk, the looks-like track removes market risk, and you sequence them according to which risk is currently larger. More on planning that sequence is on our prototyping hub.

Decide the right prototype before you spend

Projects House builds both kinds - functional units that prove the engineering, and appearance models that sell the idea - and will tell you plainly which one your project needs next rather than quoting for both. Describe your project through our contact form and we will recommend the shortest path to the answer you actually need.