Custom Dimensions in Analytics: Track Extra Context Beyond Pageviews

Standard analytics tells you that 1,200 people visited your pricing page this month. Custom dimensions tell you that 840 of them were free-plan users, 310 were on a trial, and 50 were already paying customers. Without the extra context, you're looking at aggregate traffic. With it, you're looking at behavior by segment — and that difference is where almost all useful analysis lives.

Custom dimensions (also called custom properties or custom attributes depending on the platform) are metadata values you attach to hits, events, or user profiles that wouldn't otherwise be captured. They extend your analytics beyond the built-in dimensions — URL, browser, country, referrer — to capture the things that matter specifically to your business.

What Custom Dimensions Actually Are

Every pageview and event in your analytics data has a set of properties automatically recorded: the URL, the timestamp, the browser, the device type, the country. These built-in dimensions are the same for every site because every site shares them. Custom dimensions are the properties you define for your specific situation.

There are three scopes at which custom dimensions can apply, and the scope determines how the value is stored and how it can be used in analysis:

• Hit-level dimensions attach to a single pageview or event. Example: the category of the blog post being read, or the A/B variant shown on this specific page load.
• Session-level dimensions apply to all hits in a session. Example: whether the user arrived from a paid campaign, or which version of the homepage they saw.
• User-level dimensions persist across all sessions for a given user. Example: their subscription plan, their country from profile data, or their account creation date cohort.

The Most Useful Custom Dimensions to Capture

User-level properties

For any site with user accounts, the most valuable custom dimensions are properties from the user's profile. Subscription plan ("free", "pro", "enterprise"), whether the user has completed onboarding, the account's creation date bucket (day 0–7, day 8–30, 30+), and whether the account has an active payment method are all properties that transform standard analytics into product analytics.

When you can filter your top pages report by subscription plan, you can see which features trial users explore most — which is exactly the data that drives onboarding improvements. This is the complement to what user identity tracking enables at the individual level; custom dimensions enable it at the aggregate segment level.

Content-level properties

For content-driven sites — blogs, documentation, news sites — tagging each page with structured metadata makes your content analytics dramatically more useful. Author name, content category, publish date bucket ("older than 1 year", "last 6 months"), and content type ("tutorial", "case study", "opinion") let you compare performance across meaningful groups rather than comparing individual URLs.

Instead of "these 3 URLs got 80% of traffic," you can ask "do tutorial posts perform better than case study posts across all 50 articles?" — which is the question that actually shapes your editorial calendar. This pairs naturally with content performance analytics to move beyond raw pageview counts.

Campaign and test properties

When you're running A/B tests or targeted campaigns, custom dimensions let you record which variant a visitor saw and carry that context through the session. If variant B gets 12% more sign-ups than variant A, you want to confirm that the conversion rate difference is statistically meaningful across a large enough sample — and that requires capturing the variant as a dimension, not just tracking conversions at the aggregate level.

How to Use Custom Dimensions for Segmentation

Custom dimensions only deliver value when you actively use them to filter and compare. The most common analytical workflows they enable:

• Compare behavior by plan tier. Free users vs. trial users vs. paid users almost always behave differently. If paid users visit your pricing page at half the rate of trial users, you've found something worth investigating — are they looking for an upgrade path, or checking a competitor?
• Find content that converts by type. If "how-to" articles generate signups at 3x the rate of "opinion" articles, that directly informs which content to prioritize.
• Measure onboarding effectiveness by cohort. Tag users with their account age in weeks, then compare engagement metrics across cohorts. If week-2 users visit the integrations page significantly less than week-1 users, your onboarding may not be surfacing integrations at the right moment.
• Debug campaign performance by variant. When a campaign underperforms, custom dimensions let you check whether it's the creative, the landing page, or the audience segment that's dragging down results.

The Bottom Line

Custom dimensions are the difference between analytics that describes traffic and analytics that describes your audience. The implementation requires a small amount of upfront work — deciding which properties matter for your business, pulling them from your data layer or user session, and passing them to your analytics snippet. But the analytical capability you gain from even three or four well-chosen custom dimensions is substantially larger than any number of additional built-in reports. Start with one user-level dimension (plan tier or account status) and one page-level dimension (content category), and build from there as you identify the questions your default analytics can't answer.

For hands-on event tracking, see how custom event tracking works alongside custom dimensions to build a complete picture of user behavior.