How to design a BIM family
in Revit, step by step
A BIM family is the ingredient from which any model is built: a door, a diffuser, a piece of equipment. Designing it well — with a clear concept, the right template and parameters that are understood at first glance — determines whether that family serves for years or has to be rebuilt in the next project.
Keep reading: the 6 key points for designing a family in Revit ↓
Design a concept before opening Revit
The quality of a BIM family is decided before touching the software: in how you think about the object it will represent.
A family is, after all, an object — a table, a chair, a door — that has to be designed as a set of elements, not as a single shape. That is why the first step, before creating anything in Revit, is to have a well-defined preliminary scheme: which components the object has, how they relate to one another and which parts must be able to vary depending on the project in which it is used.
The most effective approach is usually to start with a hand drawing that helps to fix those components before modelling them. Once the concept is clear, it is resolved with nested families — families inside families — to group each part of the object instead of forcing everything into a single monolithic element. That early decision is what later determines whether the family is easy to maintain or whether any change forces you to rebuild it from scratch.
A BIM family is designed first in your head — or in a hand sketch — and only then in Revit: the preliminary concept decides whether the rest of the process runs smoothly or turns into successive patches.
Choose the template according to where the element will live
The starting template is not an administrative detail: it conditions how the family behaves the day it is placed in the project.
The two templates most used to create BIM families are Metric Generic Model and face-based. The first is suited to an independent element — a piece of furniture, for example — that is placed freely in space. The second is the better choice when the element can be hosted on different planes: ceiling, wall or floor, depending on the point of the project where it is used.
A common practical case: an electrical element that normally goes on the ceiling but, in some projects, may end up on a wall. Creating that family with the face-based template allows it to be placed wherever it is needed, with no tie to a fixed orientation — the family inherits that of the plane in which it is inserted, instead of forcing the modeller to rebuild it for each case.
Work with clearly identifiable parameters
A badly named parameter is a silent trap: it works today, but becomes unreadable as soon as someone else — or you, months later — has to touch the family.
When you create the parameters that govern the visibility and dimensional properties of an element in Autodesk Revit, each one should identify itself, without depending on context to be understood. In a table, for example, it is not a good idea to use “width”, “height” and “depth” in general: it is more useful to separate each component and name its parameter accordingly — “Leg.width”, “Leg.height” — rather than a generic one that fits anything.
From there, grouping parameters by function — dimensions, visibility, materials — is what turns a family into an object that is easy to audit. The naming and grouping of parameters are the key to a BIM family being understood without having to open its documentation each time.
Test the family at three levels before considering it finished
The earlier a behavioural error is detected, the cheaper it is to correct.
As the family scheme is built, its properties should be tested at three levels. First, in the family properties editor, to confirm that no errors appear when values change. Then, in the family environment, repeating the same check outside the editor. And finally, in the real BIM project, loading the family into the model even though it is not yet finished.
That last step — loading it into the model before considering it complete — is what reveals what fails in a real context: how it relates to other elements, whether its constraints conflict with those of the project, whether the visual behaviour changes depending on the view. Correcting it at that point is far cheaper than discovering it when the family is already in production across dozens of instances.
Testing at three levels — family editor, family environment and real project — turns testing into a habit, not a final step that gets skipped when the deadline tightens.
Shared nested families: use them with judgement
They are a powerful tool for quantification, but not a free one: each level of nesting adds weight to the model.
Shared nested families are used to count elements independently within another family. For example, if you need to know how many shelves a bookcase has, marking that nested family as shared gives you that figure immediately, instead of treating the whole bookcase as an indivisible unit.
That said, working with this type of family can cause performance problems if it is overused: each level of nesting adds weight to the file and complexity to the model calculation. Except where that independent count is genuinely needed, it is preferable to reserve shared nested families for when they bring clear value, not as a default practice.
Keep a record of the different versions of each family
A BIM family is rarely perfect the first time: it evolves, and that evolution must be traceable.
There is something we do at A3D with every BIM family we produce: we keep a record of its different versions to avoid confusion when that family is implemented in different models. Without that record, it is easy for two projects to end up using different versions of the same object without anyone noticing until an inconsistency appears in the documentation.
Keeping track of each element and of the improvements made to it is, at heart, the way to build a BIM family library that is clearer, more complete and useful across all projects — not a folder of loose files with no traceability.
The rest of the series on BIM families
This is the first of four pages on how BIM families are designed, managed and applied in a real project.
- Building services Characteristics of MEP families in Revit →
- Complexity Complex or assembled families: which to choose →
- Governance BIM family management in an architecture or engineering practice →
Frequently asked questions
Complete answers to the most common searches about designing BIM families in Revit.
What is a BIM family in Revit?
A BIM family in Revit is an object — a door, a table, a piece of equipment — built as a set of elements with their own geometry and parameters, not as a loose drawing. When it is inserted into a project, the model recognises what it is, what material it is made of and which variants it supports; that is why a well-designed family is reused across many projects without being rebuilt each time.
How do you start designing a family in Revit?
You start with a well-defined concept before opening Revit: which components the object has, how they relate to one another and what must be able to vary. A hand sketch helps to fix that idea; from there it is resolved with nested families — families inside families — that group each component instead of forcing everything into a single element.
Which template is used to create a family in Revit?
The two most common templates are Metric Generic Model and face-based. Metric Generic Model suits independent elements, such as furniture; face-based is the right choice for elements that can be hosted on different planes — ceiling, wall, floor — because it inherits the orientation of the surface where it is placed, without forcing you to rebuild the family for each case.
How do you name the parameters of a BIM family properly?
Each parameter must identify itself, without depending on context to be understood: “Leg.height” or “Leg.width” is better than a generic “height” or “width” that fits anything. Grouping parameters by function — dimensions, visibility, materials — avoids ambiguous names and makes the family legible to anyone who opens it later, not only to whoever created it.
Why should a BIM family be tested before it is finished?
Because the errors of a poorly resolved family — geometry that breaks when a parameter changes, constraints that conflict — are detected earlier and corrected at lower cost if they are tested at three levels: the family properties editor, the family environment and, finally, the real BIM project. Loading the family into the model before considering it finished shows how it behaves in a real context, not only in isolation.
What are shared nested families and when should they be used?
They are families inside other families that are marked as “shared”, which allows them to be counted independently in the project — for example, knowing how many shelves a bookcase has without counting the whole bookcase as one unit. They are useful for quantification, but each level of nesting adds weight to the model: reserve them for cases where that count is genuinely needed, not as a general rule.
We do this work every day
Designing BIM families to this level of detail — concept, template, legible parameters, testing and version control — is exactly what we do at A3D every day, for our own projects and for the engineering firms and practices we work with. If your team needs a family library of its own, Revit training or targeted support so as not to have to solve it from scratch, we go through it together on your project.
And your family library: what state is it in?
Tell us about your plant: an A3D specialist tells you in 15 minutes where to start, which area to survey first and with what scope.
Prefer to continue with the series? Read about MEP families in Revit or go back to the 6 key points on this page.