Why you shouldn’t jump straight into CAD
When you’re making an idea into a physical object, it’s a logical line of thought to want to get a model of that idea as soon as possible. It feels like progress. It produces beautiful renders. You can drop shots of it into a pitch deck and watch the room think that you’re only a few steps away from the final product.
CAD models seem like the first meaningful step, and in many ways, they are. To make an object, you have to have a manufacturing plan. Whether that’s injection molding, casting, machining, or otherwise, a precise model is typically needed for most objects to be able to move from a rough prototype into a mass-produced product. CAD is where details get finalized; draft angles get added; material thicknesses get dialed.
A finalized CAD Model is one of the most satisfying things you can get out of a product development project. That’s the problem.
We’re not against CAD. Nothing happens without it in some shape or form at the end of the day. Everything that was once an idea has to go through the process to get to the shelves. The argument here is about order. Finalized CAD is an important part of the process, but it’s not the first step after the idea. It’s not the second either. CAD belongs near the end of the process. Starting with CAD doesn’t speed up a project. It just gives you more time to perfect details that may become irrelevant once your assumptions change.
Why even good teams get tricked into starting with CAD
When products jump into CAD quickly, it’s not because anyone is actively deciding to skip ideation. When a product ends up in CAD too soon, it’s typically for reasons that make perfect sense.
For instance, ideation is hard to read as a non-technical person. It’s much easier to judge a CAD Model. Napkin sketches are funny things. They might look like a scribble to someone who isn’t familiar with the context. Similarly, a rough cardboard mockup can feel like you are spending money on someone working on something akin to a middle school art project rather than a six-figure product development process. It’s easy to struggle to explain this to a client or investor in a nutshell. On the flipside, a rendered CAD model doesn’t have any surrounding questions in the same way that a pile of sketches does.
Getting to this part of the project faster feels like a relief and alleviates the pressure that mounts during the product development process. You have momentum, a release date on the calendar, and you’re trying to get there as fast as you can. This makes CAD seem like the smart move, and all the other stuff trivial.
Beyond this, if a team is trying to build multiple versions of a product, you can argue that CAD is the sketchpad. Parametric modeling allows you to iterate, change dimensions, and see what breaks as a result. That’s a completely valid use case.
None of these reasons are dumb in a vacuum. None are even entirely wrong. The issue? All of these paths assume one crucial thing: You already know what you’re building.
By stunting the product development process by jumping straight into CAD though, you don’t.
Get weird before you get serious
Frontloading the process with weird and wild ideas isn’t just a novelty. It helps teams view their problem from dozens of different angles.
Napkin sketches feel more like a metaphor than a reality oftentimes, but there is real benefit to starting the development journey at such a low-fidelity point: rapid ideation. It’s easy to dwell on the original idea. It’s the seed from which the product being developed bloomed, and there is always some level of attachment to that original idea. That attachment only grows if it’s the only idea nurtured. This leads to projects falling into echo chambers of development where rather than the best path forward being taken, the path is distorted and skewed to fit in line with the original product vision. Just because an idea got you to start the process does not mean that same idea is going to get you over the finish line.
Sketching fast breaks up these siloes of thought. You put six people into one room. Put 15 minutes on the clock, and allow their minds to all go wild in different directions with no constraints. That one idea quickly becomes dozens and dozens of different ones. They might not all be great. In fact they shouldn’t be. Early in the process when you are figuring out exactly what to build, you shouldn’t constrain yourself to a single vision. That leaves you with massive blind spots.
The sketches themselves aren’t the point of this part of the process. The point is finding out what constraints you had in your head are tangible, and which ones were just blockades you had put up to have everything fit your initial vision. By just taking that original idea straight to engineering and then sprucing it up afterwards instead, you leave potentially crucial improvements on the table.
Ideation Buys Answers Before You Buy Parts.
Once you’ve got concepts on the wall, the immediate instinct is to pick one and run to engineering to finish it up. Hold off.
Before spending the time and money on tooling or high end prototypes, it might make more sense to spend your time in a much cheaper medium that can answer similar questions. Cardboard, foam, FDM prints. All these things are rougher around the edges, but that doesn’t matter.
What’s valuable at this point in the process is answers per dollar. Is the grip too wide for a small hand? You get that answer from cardboard in an hour. A machined prototype tells you the exact same thing three weeks and a few thousand dollars later. Is a button in an awkward spot once someone is actually holding the thing the way people hold it? You can use a 3D Printer, make four versions overnight and hand them out to people around you. Quick and dirty testing provides the same value that that detailed CAD would.
It’s also important to note that quick CAD and detailed CAD are two very different beasts. Volumetric studies, rough geometry, and quick mechanism checks aren’t the same thing as that final CAD model that looks pretty. It’s faster, cheaper, and leaner. Using CAD like this to focus on quick answers is a great tool early on. Using it to build out the entire thing at once is not.
The failure we see most often is teams who skip this step and end up learning the same lessons from their first factory samples rather than a twenty minute mockup. By then the tool is cut or the cost of a prototyping house is done and the schedule is set. So the fix stops being a decision and starts being a compromise.
Design and Engineering should work together, not independently.
Beyond this, when engineering CAD is the main way that projects move forward, you default to a scenario where engineering ends up making design calls. Not because the engineers want to make those calls, but because someone has to make the decisions, and if you’re jumping into CAD immediately, that’s who’s going to make them more often than not. Designers will have input for sure, but secondary input with two pre-done choices versus wide range discussions about what the mechanism will be, where it will go, and how it will be packaged are two very different things.
Engineering’s job is to make things buildable, reliable, and efficient to produce. Conflating those responsibilities with Design’s job, which is to make a thing make sense in the hands of the person who is holding it, can kill projects.
This can go both ways. A design-led product with nobody pushing back from an engineering perspective can, if not careful, turn into a manufacturing problem. Surfaces that can’t be molded without a fight or three unnecessary slides. Parts that lead to extra time and money on the assembly line. A part count that quietly eats the margin.
Neither group is at its best on its own. By starting with CAD immediately, you exponentially increase the chances of this set up happening.
Low-fidelity work keeps both people in the room at the same time instead of one after the other. When the thing on the table is a foam block, or a volumetric model produced by rough, quick CAD, neither side owns it. An engineer can say that proportion won’t fit the board before anyone’s in love with it. A designer can push on how it feels while changing it is still free. You get a real back-and-forth instead of a handoff followed by an argument.
A product can be perfectly engineered to solve a problem. But if that problem doesn’t need solving, that doesn’t matter. Starting with a wide lens and narrowing down throughout the process makes sure questioning the why of a product is understood early on, rather than after the project is already dead, having failed on the market.
So, when is the right time to hop into detailed CAD?
Once ideas have had their chance to be explored, prototyped, tested, and refined, THEN comes time for refined manufacturing CAD.
By allowing the process time to develop, there aren’t 18 rounds of final CAD (although there might be a few, product creation shifts and flows up until the final buzzer). Here are some questions that need to be answered in practice before it’s time for the final CAD model.
What is the problem being solved? What is the current solution, and how does this design substantially improve it? Are we leaving any solutions on the table?
Is there any other form factor that we can look at that might reduce overall number of parts/reduce tooling costs?
Is the layout settled? Major parts, rough sizes, how it opens, how it goes together?
Is there any question left that a cheap mockup would answer faster than a CAD model?
When questions such as those are settled, then an engineer can follow a clear path to create the final product. It’s important to still note though:
It’s not a straight line
It’s easy to assume that at any point of the process, you just need one sketch, or one model, and you’re good to go. The better way to picture it is a road that gets wide and then narrow, over and over. Wide when you’re pulling in options. Narrow when you commit to one. Wide again when a test shows you something you didn’t see coming. And technically, you can run your project on a single, narrow path. You’ll just be hit with delays and problems in later stages when it costs more money and time than it would have if you just did the work up front.