Conceptual design is the earliest phase of the engineering design process, where a team defines the core idea for a product or system and explores possible solutions before committing to a specific direction. It sits between the initial requirements (what the product needs to do) and the more detailed work of sizing components and selecting materials. Despite being the phase where the least money is spent, the decisions made here carry enormous weight. NASA estimates that while only about 15% of total costs are spent during design, the design itself commits roughly 75% of a product’s entire lifecycle costs. Other analyses put it even more starkly: conceptual design alone determines over 50% of a product’s final cost and quality.
What Happens During Conceptual Design
The phase begins once the engineering team has a clear set of requirements, meaning they know what the product must accomplish, who will use it, and what constraints exist. From there, the work follows a general sequence: clarify and break down the problem, search for possible solutions, analyze and refine those solutions, then evaluate them to pick a winner.
In practice, this means a team might spend weeks sketching alternative approaches to the same problem. If you’re designing a new aircraft, this is the stage where you decide whether it’s a turboprop or a jet, what its mission profile looks like, what materials and avionics technologies are on the table, and how it will be manufactured and maintained. If you’re designing a consumer product, this is where you choose between fundamentally different mechanisms for accomplishing the same task. The goal isn’t to nail down every dimension and bolt. It’s to land on a solution principle: the basic approach that will guide all the detailed work that follows.
A good concept coming out of this phase needs to pass three tests. It must be logically feasible (the physics and logic hold up), functionally simple (it doesn’t introduce unnecessary complexity), and physically certain (it can actually be built with available technology and materials).
Functional Decomposition
One of the core techniques in conceptual design is functional decomposition, which means breaking a complex system down into its individual functions until each piece does exactly one thing. The classic test: if you need the words “and” or “or” to describe what a block in your diagram does, it hasn’t been decomposed far enough.
The decomposition starts at the top with the system’s overall purpose, then splits into main functions, then subfunctions, continuing until you reach basic operations that can’t meaningfully be divided further. Engineers working in this framework typically categorize functions into five elemental types: channeling, connecting, varying, changing, and storing. These apply across three kinds of flows: material, energy, and signals. That gives you a working vocabulary of about 15 basic functions that can describe what’s happening at the lowest level of almost any engineered system. The value of this exercise is that it forces the team to think about what the product needs to do before jumping to how it should look.
Generating and Selecting Concepts
With functions mapped out, the team moves to brainstorming, ideally with six to ten people contributing ideas. The point is volume: generate as many possible solutions as you can before filtering. Some concepts will be wild and impractical, and that’s fine at this stage. The creative breadth is what ensures you don’t overlook a strong approach simply because a more obvious one came to mind first.
Narrowing down happens in three steps. First, screening eliminates concepts that simply aren’t feasible, whether for physics, cost, safety, or manufacturing reasons. Second, the surviving concepts are compared against a set of evaluation criteria. Third, the team makes a final selection.
The criteria used for comparison typically fall into five categories: relevance (does it address the actual need?), efficiency (does it use resources well?), effectiveness (how well does it perform?), impacts (what are the broader consequences?), and sustainability (can it be maintained long-term?). Each criterion gets weighted based on its importance to the project, so a medical device might weight safety far above aesthetics, while a consumer electronics product might weight manufacturability and cost more heavily.
The Decision Matrix
A common tool for making the final call is the Pugh matrix, sometimes called a decision matrix. You list your candidate concepts along one axis and your weighted criteria along the other. One concept, or the current product if one exists, serves as a baseline scored at zero. Every other concept is rated against it: better (+1), same (0), or worse (-1) for each criterion. Multiply ratings by weights, sum the columns, and you get a score for each concept. The highest score doesn’t automatically win, but the numbers give the team a structured foundation for discussion rather than relying on gut feeling or the loudest voice in the room.
What the Phase Produces
Conceptual design generates a range of deliverables that vary by discipline but share a common theme: they’re detailed enough to communicate the idea and estimate costs, but not so detailed that they lock in specifics prematurely. For a building project, the University of Michigan’s engineering guidelines list typical deliverables including a scope of work narrative, preliminary site plans, structural scheme plans, typical floor plans and elevations, one-line diagrams for mechanical and electrical systems, a landscaping concept, a preliminary finish schedule, sustainability features, renderings, and a preliminary cost estimate.
For a product design effort, the deliverables might instead be functional diagrams, concept sketches, a written description of the solution principle, rough performance estimates, and an initial cost model. The key output across all disciplines is that the team has committed to a direction and can justify that choice with documented evaluation.
Where Conceptual Design Ends
The boundary between conceptual design and the next phase, called embodiment design, is genuinely blurry. Some frameworks, like Stuart Pugh’s, merge the two stages entirely. In most models, though, the distinction works like this: conceptual design picks the approach, and embodiment design gives it physical form.
During conceptual design, all options are open. You’re choosing between fundamentally different ways of solving the problem, and you’re considering a large range of possible materials and configurations. Once you move into embodiment, you take the winning concept and start sizing components, selecting specific materials from a shortlist, analyzing performance at an approximate level, and producing a preliminary layout. One useful analogy from engineering educator George Dieter: if the concept is the skeleton, embodiment design is putting meat on the bones. The embodiment stage ends with a feasible layout that feeds into detailed design, where exact dimensions, tolerances, and manufacturing specifications are finalized.
Software Tools for Concept Generation
Traditionally, conceptual design relied on hand sketches, physical models, and structured brainstorming sessions. That’s still common, but newer tools are changing what’s possible early in the process. Generative design software can now produce multiple geometry options based on constraints you define, like load requirements, material choices, and manufacturing methods.
More recently, AI-powered platforms have pushed this further. Some tools accept a text description of a part and generate a 3D CAD model directly, no manual modeling required. Others let you sketch a rough concept and refine it through guided prompts. These tools are most useful for accelerating the exploration phase, helping teams visualize and compare more options in less time. They don’t replace the engineering judgment needed to evaluate those options, but they lower the barrier to generating them.
Why This Phase Matters Most
Engineers sometimes call conceptual design the phase where you have the most freedom and the most leverage. Early on, changes are cheap. You’re working with sketches, diagrams, and rough calculations. Once you move into detailed design and manufacturing, changing direction becomes exponentially more expensive. A poor concept that makes it past this gate will create complexity downstream in manufacturing, maintenance, and operation that no amount of detailed engineering can fully fix. That asymmetry, between how little is spent and how much is determined, is why experienced teams invest heavily in getting this phase right rather than rushing to start building.