One of the most common mistakes in product development is assuming the next step is simply to “get a CAD model made”.
CAD is a fantastic tool. It allows designers to explore, define and develop physical products in a digital environment before they exist in the real world. But CAD is not a magic starting point. Used at the wrong time, or in the wrong level of detail, it can make a weak idea look more developed than it really is.
Many product projects do not go wrong during CAD. They go wrong before CAD starts.
They go wrong because the user need is unclear. The manufacturing route has not been considered. The target price has not been explored. The prototype has no clear purpose. Or the founder has not yet worked out what the product needs to achieve for them commercially.
Before you invest heavily in detailed design work, it is worth stepping back and asking a better question:
What are we actually trying to prove?
When people talk about CAD, they often talk about it as if it is one fixed output.
In reality, CAD can do several different jobs during a product development project.
At Mouse Design, we may use CAD to create early block models. These are simple, open-ended models that help explore the size, layout and physical space a product may occupy. They are not finished designs and they are not usually suitable for manufacture.
We may then use CAD to develop concept models. These help visualise the idea more clearly and explore how different design directions could work.
Later, we may create prototype-ready CAD. This is used to produce test parts, often through 3D printing, CNC machining or another prototype process.
Only once the design direction, function, materials and manufacturing route are properly understood does it make sense to move into detailed engineering CAD for manufacture.
The problem comes when people expect the first CAD model to do all of these jobs at once.
A CAD model can make a product feel real very quickly.
That can be useful, especially if you need to communicate an idea to investors, colleagues or customers. But it can also create false confidence. A product may look convincing on screen but still be completely unresolved from a manufacturing point of view. A 3D printed prototype may look good in your hand but still be unsuitable for injection moulding, sewn construction, assembly, tooling or volume production.
For example, if a plastic product is eventually going to be injection moulded, the CAD needs to consider things such as wall thickness, draft angles, tool split lines, material behaviour, assembly, screw bosses, ribs, tolerances and ejector pin marks. If those things are ignored early on, the product may need to be heavily redesigned later.
That is why the question should not be “can you make me a CAD model?” A better question is: What type of CAD model do we need at this stage, and what decision is it helping us make?
Before we start designing a product, we need to understand why it exists. That sounds obvious, but it is often where early-stage projects are weakest.
A product being developed for a lifestyle business may need a very different development route from a product being developed for seed funding. A side-hustle product may need different risk levels from a fully funded startup. A product being developed for grant funding may need different evidence from a product being prepared for retail launch.
At Mouse Design, we want to understand the person, the business and the reason behind the idea before we put pen to paper. What does success look like? Is the goal to raise money? Launch a small batch? Create a second income? Prove a concept? Improve an existing product? Build something that can scale?
Good product design starts with context.
Without that context, the design team can create something that looks good but does not actually move the client closer to the right outcome.
Another common mistake is designing the product before thinking about how it could be made. Manufacturing route has a huge influence on product design. A product that will be 3D printed, injection moulded, sewn, fabricated, vacuum formed, CNC machined or assembled from bought-in parts will need to be approached differently.
The same idea may need a different design depending on likely production volume, target selling price, material choice, tooling budget and route to market.
For example, a startup may only need a convincing prototype for testing or funding. An established brand may need a design that can go into reliable production at scale. Those are different problems, even if the product idea sounds similar.
If manufacturing is considered too late, the design can become expensive to change. The further down the road you get, the harder it becomes to turn around.
This is why early feasibility work is so valuable. It helps set the compass bearing before too much time and money has been committed.
A prototype should not exist just because “we need a prototype”.
A prototype should answer a question.
Different questions need different prototypes. A rough block model may be enough to test size. A visual model may be enough for a funding conversation. A functional prototype may be needed to test a mechanism. A sewn sample may be needed to check construction and usability.
The mistake is spending too much money on the wrong prototype too early. Before CAD or prototyping starts, it is worth deciding what you actually need to prove.
A product idea can be interesting, useful and technically possible, but still not make commercial sense. That is why target price, margin and market position need to be considered early. If the product needs to retail at a certain price, that has implications for materials, manufacturing process, complexity, assembly, packaging and production volume. If the idea requires expensive tooling, but the market is small, that may affect whether the development route is sensible.
This does not mean every cost needs to be known at the start. But the commercial direction needs to be understood enough to guide the design.
Designing a product without thinking about cost is one of the fastest ways to create something that is attractive but commercially difficult.
At the start of a project, everything feels flexible. But as design decisions are made, the project becomes more fixed.
Once detailed CAD, tooling, prototypes or supplier conversations begin, changing direction becomes more expensive. That is why a small amount of thinking at the start can save a significant amount of time later.
A product feasibility review can help identify:
It is a way of slowing down at the start so the project can move faster and more confidently later.
We recently worked with a client who had an idea for a sports training product. The initial request was around producing drawings and prototypes.
Rather than jumping straight into detailed design, we started with a feasibility review. This allowed us to understand the product idea, the user context, the development risks and the best route for testing.
The outcome was a clearer plan that helped the client move into structured testing with relevant sports clubs. That meant they could start proving the idea with real users before committing too much time and money to a full route to manufacture.
In that case, the most valuable first step was not CAD. It was understanding what the product needed to prove.
Before you commission detailed CAD, it is worth asking:
If you can answer those questions, CAD becomes much more powerful. Instead of simply creating a digital model, it becomes part of a clear development process.
CAD is an important part of product design, but it is not always the right place to start.
The early thinking around user need, business goals, manufacturing route, cost, prototype strategy and feasibility is often where a project is won or lost.
At Mouse Design, we help clients set the right direction before they commit to full development. Sometimes that leads to CAD. Sometimes it leads to a design sprint, prototype test, manufacturing review or a completely different route.
The important thing is knowing what you are trying to prove before you start building the wrong thing in detail.
