Unlike traditional CAD (Computer-Aided Design) methodologies, where objects are not interrelated and revisions require changes to be applied mostly manually, parametric systems create networks of dependency among different design objects, allowing changes to be made almost automatically before a line is even drawn.

The traditional approach essentially merges the drawing process with a drafting table, creating three steps:

1. Apply the command
2. Define parameters
3. Accept, change, or erase

If we decide to change the design intent, we must essentially start over and repeat the commands until we achieve the desired result.

Figure: Production in traditional CAD systems

This stands at a clear disadvantage in front of parametric systems which allow us to express a design intention and explore multiple options before actually committing to a solution.

In a parametric system, the algorithm exists “isolated” from the drafting program. This allows us to explore various possible configurations (theoretically infinite) before translating them to the program. This effectively reduces the potential work to two steps:

1. Designing the algorithm
2. Accepting or changing the result

In a parametric system, erasing is not necessary because the script exists outside the drafting program until the result is summoned. Additionally, the information remains reusable for other projects with similar configurations.

Figure: Production in typical parametric systems

This gives designers much more time to focus on finding the optimal solution, either by quickly producing different comparable alternatives or by allowing for exploration that can be directly translated into results. Solutions generated this way have a high degree of adaptability, enabling the generation of endless possibilities given a sufficiently large data set and a comprehensive algorithm.

Parameters are not only restricted to generating complex geometrical forms in design applications; the ability of computer systems to read and analyze information allows for the relative automation of many design processes that might be too complex or tedious otherwise.