Although technically we can model almost anything, in reality, the process is often limited by a lack of experience, time, and resources. This can lead to inefficient workflows, inaccurate results, and information that’s simply not usable in practice. That’s why the first question to ask before starting any modeling is simple: what will the information be used for? This answer helps define what needs to be modeled, when, to what level of detail, and what information should be included.

For buildings, it’s generally recommended to model all elements included in the construction design. This allows for data integration, quantity take-offs, clash detection, and visual evaluation of design solutions.

The main exception is electrical cables - modeling them takes a lot of time and experience. In my own house project, inaccurate cable quantities turned out to be one of the most expensive mistakes in the cost estimate, so investing a few hundred, in modelling cables, could have saved me a few thousands.

For external infrastructure, again, it depends on the purpose. If you need to visually evaluate solutions (e.g., building placement, landscaping), modeling can be very useful. But if your goal is simply to get accurate quantities, 2D drawings, such as plans and lines, are often enough.

Likewise, it’s not worth investing too much time in clash detection if you lack reliable data on existing networks or if resolving those clashes won’t bring any real added value.

When requesting a 3D model, it’s essential to define its purpose from the start.

Don’t confuse modeling with detail - having a model doesn’t mean it's detailed or should be. Sometimes a simple box is enough to understand that a proposed solution will cause problems on site.

Avoid making major design decisions in the later stages of the project, especially if they impact multiple disciplines (like spatial layout), because making changes to complex models can be time-consuming and lead to unintended consequences.