Custom Dimensions let you group data from multiple sources — Shopify, Meta, Google Ads, and more — into a single, unified breakdown in your reports. This guide walks you through building one from scratch and fixing common issues when data isn't grouping the way you expect.

When you connect multiple platforms to Polar, each one stores similar information under different names. Your Shopify store, your Meta ad account, and your Google Ads account all have a concept of "country" — but they call it Shipping Country, Country, and Target Country respectively.

Without a Custom Dimension, pulling all three into one report gives you fragmented rows that can't be compared side by side. A Custom Dimension fixes this by letting you define the grouping yourself: "anything with Shipping Country = United States or Target Country = United States should appear as the United States row in my report."

Each Custom Dimension is built using a When / Then / Else structure. You define the criteria a data point must meet (When), what value to return when those criteria are met (Then), and what to return when nothing matches (Else). You can stack multiple When/Then blocks to create as many groupings as you need.

You can launch the Custom Dimension builder here, or in any of these three locations:

Your When statement defines the criteria that must be met for the dimension to return your desired value. Each When statement requires three components, which you select from the dropdown:

To add multiple conditions to the same When statement, use the OR or AND options that appear to the right of your first condition.

Your Then statement defines the value Polar should return when the When conditions are met.

In this example, we added a string value of "North America" by typing this into the "Then" bar.

Add more When/Then blocks as needed To create additional groupings (e.g. a separate group for Canada, Europe, etc.), add another When/Then block for each one. Polar evaluates them top to bottom — order matters when rules could overlap. See Section 5 for details.

Your Else statement defines what Polar returns when none of the When conditions are met — this is your catch-all. Like the Then statement, you can enter a string value (e.g. Rest of the World) or return an existing dimension with a function.

In this example, we added a string value of "Rest of the World" - this will ensure that all data points that do not have the United States as their Shipping Country are defined as "Rest of the World".

Once your Else statement is set, the"Which data are selected in your dimension" preview appears at the bottom of the screen. This shows you exactly how your data will be grouped and updates every time you change a rule — use it to validate your logic before saving.

Give your Custom Dimension a clear, descriptive name (e.g. Target Region or Brand) — this is what appears in the breakdown and filter pickers across the app. Then save to make it available everywhere in Polar.

Operators define how a dimension value is compared against a condition.

Polar evaluates your buckets in order from top to bottom. As soon as a data point matches a bucket's rule, it's assigned to that bucket and the evaluation stops — it won't be checked against any remaining buckets below.

This matters when your rules could overlap. A common example: you have a Shopify store for Europe that you use to bucket all European sales, but you also want Germany to appear as its own separate bucket.

In the Report Builder, you can drag and drop buckets to reorder them. Put the most specific rules at the top and broader catch-all buckets toward the bottom.

Any data point that doesn't match any of your defined buckets automatically falls into the Other(or Else) category. This is useful as a catch-all, but if Other contains significant volume, it's a sign that some values in your data aren't covered by your current rules. See the Troubleshooting section for how to track these down.

A Custom Dimension will only be available to select in a report if the metric you've added has a source that's included in the dimension's mapping. If you add a metric from a platform that isn't mapped in the dimension, the dimension won't appear as a breakdown option.

Don't wait until your dimension is fully built to check it. Validate one bucket at a time — it's much easier to catch and fix a mismatch early than to untangle everything at the end.
​
​Quick validation steps

If your dimension isn't grouping data the way you expect — too much is in Other, a bucket has less than it should, or a platform seems to be missing entirely — work through these steps to find the cause.

If you build a Custom Dimension using Shopify tags — like order tags, product tags, or customer tags — you may notice that your row totals appear to add up to more than the overall total. This is expected behavior and not a bug, but it's important to understand why it happens and how to avoid it.

Polar stores tag data as individual line items. An order can have multiple tags applied to it at the same time. If you use the singular Order Tag dimension to group data, each tag on an order creates a separate row — meaning the same order's revenue can appear in multiple rows simultaneously, inflating your totals.

Instead of using Order Tag, use Order Tags in Order (note the plural form). This dimension checks the full tag list on the order as a single value, rather than breaking each tag into its own row. Combined with the Contains operator, this lets you match a specific tag without creating duplicate rows.

You want to group orders tagged wholesale into their own bucket.

❌ Wrong: Source dimension = Order Tag, operator = Is, value = wholesale

This will count a single order multiple times if it has more than one tag.

✅ Correct: Source dimension = Order Tags in Order, operator = Contains, value = wholesale

This checks whether the order's full tag list contains "wholesale" and counts the order exactly once.

The same logic applies to Product Tags and Customer Tags — always use the plural, "in order" variant when filtering by tags.\