How to have the best SketchUp models ever

If you want to know the keys to a good SketchUp model without reading a book about it, you’ve come to the right place. Admittedly I am occasionally more long-winded than a book, but it comes from experience that i wish to share, in order that your efforts with SketchUp are more effective and efficient. Even after using it for years, if you are missing one of the fundamental keys to SketchUp, it could be keeping you from a truly useful model. I realize others may work differently from me; this is just what I've found to be best for how I use SketchUp everyday. 

Let’s dive in. 


Screenshot 2014-01-06 01.27.38.png

Layers in SU are a way to organize your model. They can help control the visibility of geometry as you model, and for what you want showing at a given time in the final model. The benefits of using Layers are many, but my advise deals with this single fundamental of Layers in SU: "Draw everything on Layer 0. Always.” 

In my workflow all raw geometry, without exception, lives on Layer 0. (Raw geometry is just whatever you draw in SU that’s not yet contained in a group or component). But why, you ask, would I limit myself to Layer 0 when i have potentially a metric ton of layers? Because the layers were never intended for raw geometry. Let’s consider beans.

  Beans are like raw geometry

Beans are like raw geometry

If you were to purchase beans from a market you would likely find them on shelves in packages. Or contained in a bulk bin. I’ve never seen beans just sitting on shelves here and there, unless a package was ripped open and the beans were spilled. If that happens you know that as you grab something from the shelf you get showered with the loose beans that are helpless but to come along for the ride with whatever they were resting up against before you move it. Even when you get home with the beans, you don’t pour them onto your shelves- navy beans on this shelf, white beans on that shelf, etc. You put them into containers first. You put the containers on the shelves. It’s the container that you put on the shelves- they happen to contain beans, or rice, or flour, or any number of things you wouldn’t care to store directly on a shelf.  

This brings us back to SU and layers. If you haven’t guessed, the beans are raw geometry. If you are going to organize the geometry in your model using layers, you have to first put that geometry in a container such as a group or component. Then you put that group or component on its layer. 

So why not just draw things on separate layers like we did with Autocad? (Walls on the wall layer, windows on the window layer, text on the annotation layer, etc.) The answer can be found in the next key to a truly useful SU model.


Just like the loose beans are better off in a bag or jar, loose geometry is better off contained within Groups or Components. All too often I see models in which some part of the model (if not the entire model), is just drawn as one piece of connected geometry. There’s no way to control a model that’s been built this way. What if i want to select just the roof or only the railings, maybe single out the steps, or grab just windows? What if all your beans were mixed up and you needed to quickly get a cup of just lima beans? Separating things into containers is a good place to start.

How do you know what geometry should get grouped? Start with big stuff that you might want to move around or scale or hide, and group it. Geometry that you might need to select quickly or edit later is a good candidate for groups, as are often those parts of a model that need to be defined by a specific material. Take for example a double hung window: I would give the frame, and the top and bottom sashes their own groups because at some point i might want to edit the model to show the window as opened. Furthermore I would group the glass separately because it’d be easy to quickly select the glass for editing later. And of course if I plan to ever re-use the window I would contain all four groups into a window Component. In this way if i end up with 50 of these windows in my model and decide they should all be 1” narrower and painted yellow, all i need to do is edit one of the windows components and the changes are instantly reflected in the other 49 window components!  

  Group your beans before you put them on the shelf

Group your beans before you put them on the shelf

We’re now ready to discuss why one doesn’t simply draw door geometry on a door layer, and wall geometry on a wall layer, etc. For one reason, it’s easier to draw on Layer 0 than to constantly be  switching layers every time you draw part of a door or window or stair or car, etc. And if you don’t think that’s a big deal, see what happens when you realize you’ve been drawing on the wrong layer for 10 minutes. And then what happens when you go to move geometry that’s sharing the same object but is on different layers. And then what happens when you hide that geometry! It’s not only confusing, it leaves you helpless to hide or move certain objects without adversely affecting other parts of your model. By keeping everything on Layer 0 and using Groups and Components properly you avoid a world of pain. 

I will extend this discussion just enough to caution against two things. First, avoid placing geometry in the same plane. If you draw a table surface and group it and then draw a piece of paper laying flat on that table in the same plane there will be a conflict. Simply move the paper up over the surface of the table by a fraction of an inch and, so long as the geometry is no longer in the same plane, all is well. 

The second thing to avoid is burying groups within groups within groups within components within groups within components… and so forth. You can embed groups and components within each other, just as you can stack a jar inside another jar or stuff a plastic bag full of other plastic bags. But you know that you can’t stack jars too high before the stack is unstable, or stuff too many bags before it’s too impossible to grab a single bag. My point is, keep the grouping light; avoid burying geometry, groups and components too deeply


Practically speaking, a piece of paper has two sides. We’ll call one side the front face and the other the back face. In SU, a surface is like a piece of paper- infinitely thin and having two sides. The front side is what you want facing outward for the world to see. An effective SU model is one in which all the surfaces are facing in the right direction. This can be checked very easily by viewing the model using the Monochrome face style. By default the back side of geometry is purple and the front is white (depending on your style template). 

  To the far right is the Face Style: Monochrome

To the far right is the Face Style: Monochrome

In SU as you draw and push/pull, geometry sometimes ends up facing the wrong direction. You might find purple on a random wall in your house model, or all the glass in your window is purple, or the entire tree model you just imported is purple. Why not just use the Bucket tool and paint the purple surface white? Because the surface is still facing the wrong way. And if, for example, you render the model using Vray or Thea, or export it for use in 3dsMax, the back-facing geometry may very well render as black, or disappear all together. There are some great extensions for correcting large areas of your model with back-facing geometry, but by just right-clicking these offending surfaces and selecting Reverse Face as you model you will keep your model useful and effective. 

  Purple faces are back-facing

Purple faces are back-facing



SketchUp allows you to apply color and texture to geometry in several ways, each having its own advantage. Although this tip is just as important if you never plan to render outside of SU, it’s particularly important for those of us who render using Vray, Thea, or any of other available packages. And it is practically a must if you plan to import the SU model to 3dsMax or similar software. The fundamental is this: apply textures and color to raw SketchUp geometry and not the group or components that contains it

This is a hard one. As you go throughout your SU model applying and tweaking textures and colors it’s so very easy to end up applying textures sometimes to geometry and sometimes to the group or components. The trick is to be consistent and aware to make sure it's always the geometry and NOT the container. 

Let’s look at a glass jar of white beans. If you want those beans to be colored red, you could go ahead and color each one individually, or you could just paint the jar red so that they all appear red when viewed through the red glass. But painting the glass hasn’t changed the color of the beans themselves. Only by painting the beans themselves with the beans be made red.

This is a subject deserving it’s own essay so I won’t go into more explanation. My models are cleaner and no longer confusing to edit since I have made it a habit to texture the geometry and let the groups and components alone. 



This one is easy. When you’re ready to hand off a model, first make sure you purge from it all the extra geometry, layers, materials, and styles that have stayed with the model but are no longer needed. Purging get’s rid of all that stuff that you don’t need and makes your file smaller.

So that’s it. Your SketchUp models will be more useful to you and those to whom you may be handing them off. You will have easier access to editing your models, and you will save hours of prep work for rendering. 


Next time I’ll talk about some ways you can improve your workflow.