One of Funnel's core concepts is that of dimensions. This article will teach how to work with dimensions but will not cover the technicalities of creating them. To learn more about creation see our article on Creating your first dimension.

One misconception we sometimes notice is that our users try to create a dimension to filter out just some of their data.

Think of a dimension not as a way to filter your data, but as a way to group it.

If you want to filter out certain data you can do that by including or excluding certain dimension values. Let's look at an example.

Say you want to compare results from your advertising campaigns, but only campaigns from Facebook. Do this with two separate dimensions:

List your data by "Campaign" and apply a filter to only include data that matches "Traffic source" = Facebook.

You could, of course, also solve this by creating a dimension called "Facebook campaigns" and have each dimension value equal a campaign on Facebook. 

However, if you run ads on more than one advertising platform you probably would like a way to compare also your campaigns from other traffic sources. With the "Facebook campaigns" approach you'd have to create one dimension for each traffic source + campaign combination you might be interested in which would be a lot of work, and also limit you in other analyses you might want to do. 

If you advertise different product categories and use the naming of your campaigns to determine what category is advertised, you can create a dimension in Funnel that groups your data by product category. 

Say you want to analyse what traffic sources are performing best for a certain product category. You can do this in the same way as when we isolated campaigns that had "Facebook" as value for the dimension Traffic source.

Instead of having your data grouped into the tw dimensions "Traffic source" and "Product category" you could, of course, create a lot more dimensions to achieve the same result. You could group your data so that you have one product category grouping per traffic source:

This is completely unnecessary in most cases. If you find yourself creating a lot of dimensions in your Funnel Workspace it's probably a good idea to stop for a minute and think about what you're trying to achieve.

A good practice is to have each dimension describe one aspect of your data. "Google product categories" would fail this test since for each dimension value it describes both the advertising platform AND the product category. 

The same is of course true for our example "Facebook campaigns".  Each value in this dimension describes both the advertising platform and the campaign name.

By having each dimension describe one, and only one, aspect of your data you'll find yourself with significantly fewer dimensions to set up, as well as dramatically easier task doing your analyses.