Filter Level Counts in Segmentation

Anyone who has built segments knows the drill: add a filter, run the segment, wait for the population count, tweak the logic, and run it again. This process can take a few iterations before everything looks right.

Filter-level counts make it easier by showing the impact of each filter instantly. Instead of waiting for the entire segment to run, you can see how each filter or container affects your audience size. This gives you faster feedback, clearer insights, and far more confidence in your segment logic.

Note: For filters on related attributes, the population count is calculated at the container level rather than for individual filters within that container. For simplicity in this blog, container-level counts are referred to as filter-level counts.

Why Filter-Level Counts Matter

Segmentation is often exploratory. Marketers test different conditions until the audience looks right. Without visibility into individual filters counts, this process becomes trial and error.

Filter-level counts make this process far more transparent by showing the population impact of each filter or container.

This helps you:

• Reduce Guesswork – Instantly see whether a filter is too restrictive or barely affecting the audience.
• Build Segments Faster – Refine logic without repeatedly running full segment counts.
• Save Platform Resources – Avoid unnecessary segment executions that consume compute and credits
• Debug Segments Easily – Quickly identify the exact filter that is shrinking or inflating your audience.

Think of these granular counts as a magnifying glass for your segmentation logic.

A Simple Example

Imagine you’re building a segment for a running shoe promotion.

Your segment includes these filters joined by “AND” logic:

• Customers who purchased at least two sports apparel in the last 6 months
• Customers who engaged with a running-related campaign at least once
• Customers located in California

When you calculate filter-level counts, you might see:

• Sports apparel buyers → 120,000 customers
• Running campaign engagement → 35,000 customers
• California customers → 3,200 customers

Now it’s immediately clear that the location filter is dramatically reducing the audience. Instead of guessing why your segment size dropped, you can quickly decide whether to expand the geography or adjust the logic.

This kind of visibility helps marketers refine audiences much faster.

How It Works

Using filter-level counts is straightforward.

1. Open Segment Builder – Create or edit your segment.
2. Add Filters or Containers – Define filter conditions such as purchase history, engagement, or location.
3. Calculate Population – Click Calculate Population for any filter or container.

The system will display one of three statuses:

• Calculating Filter/Container Population – Count in progress
• Filter/Container Population – Count successfully calculated
• Population Count Failed – An error occurred

Once calculated, you can instantly see how each condition affects the segment population.

Here’s a demo you can look at which explains how to use these counts, and how they work.

When Do Counts Reset?

Filter-level counts reset when:

• The segment is refreshed
• You exit the Segment Builder
• You edit the filter or container

You can simply recalculate to see the updated population.

What about credits consumed?

• Filter-level counts are ad-hoc counts, meaning they’re calculated on-demand. This helps save you credits (if you’re on a consumption model) since we won’t always display the count unless you trigger it.
• Credit consumption for filter-level counts works just like segment counts. If you’re charged for segment counts, you’ll also be charged for container counts.
• Filter-level counts provide the population for each container, and all objects used to build that container or filter will contribute to the credit calculation.

A Note on Dynamic and Approximate Counts

Filter-level counts are not available for dynamic segments with parameterised values enabled. If Approximate Count is enabled, the system will return estimated populations instead of exact numbers. This is useful when working with very large datasets and needing quick insights.

Key Takeaway

Filter-level counts bring much-needed visibility to segmentation. They help marketers understand filter impact, debug segments quickly, reduce unnecessary runs, and build audiences faster. Instead of guessing why a segment looks the way it does, you can now see exactly what’s happening inside your segmentation logic.

Have feedback? Reach Out

In case you have feedback about this feature, please feel free to reach out to me on Linkedin @darshna sharma. I’d be happy to hear your feedback and improve the feature further.

Written by