Custom metrics are like custom dimensions (see the previous post), designed to let people tailor their GA4 setup more to their org. The difference is that custom metrics are numbers that would be added up (or otherwise aggregated) across multiple events, sessions or users. Whereas dimensions can’t be aggregated since you can’t add up (say) the text values of “Bronze”, “Silver” and “Gold”. So while dimensions would represent rows in a table, metrics would typically represent columns.

Creating custom metrics is similar to the process for custom dimensions:

• Make sure the value is being sent as a parameter. You should make sure it’s being sent as a number as not as text, some setups and plugins can send something like “2.5” which is interpreted as text which can cause some issues.
• Register the custom metric in the GA4 interface. You will need to specify whether your number represents currency (eg. cost/revenue) distance, time or if it’s just a number.

The usage of custom metrics is also very similar to custom dimensions so the main question is which custom metrics would it actually make sense to set up for your business? While almost all orgs will have use for custom dimensions, there may be cases where you don’t need any custom metrics. Here are some we’ve found before which you can use as a starting point:

• Page loading time: This can help analyse what pages took the longest to load and whether that’s impacting conversion.
• Estimated lifetime revenue: If you have a subscription product or take regular donations, it’s common to set up an ecommerce purchase event representing the first instance of the subscription. But you may have data on how long people stay for which is worth sending to compare channels across estimated LTV. You wouldn’t want to mess around with the standard revenue metric though so a custom metric would be best.
• Number of seats/passengers/covers: For booking-based businesses like travel, restaurants etc, you might want to know the number of covers/passengers that are being brought from different channels and see that across your reports.
• Cost of goods/services and net profit: You can get more sophisticated if you want to and start to model the actual profit when someone becomes a customer, not just the revenue. Useful if your margin differs a lot by product/service and if you can capture those values at the time the user converts on the website.
• Points: If you’re running an online game with points, that might be a good overall representation of what parts of the game players are finding most valuable. This can also be useful for SaaS, if you have any internal metric that represents a loyal user (or what you want your customers to do more of), it’s generally worth seeing if you can make that a custom metric in order to start seeing it in all your reports.