Building a segment, step by step

Even before building anything yourself, you'll notice a few segments already exist: Suppressed or Unsubscribed, Push Subscribers, WhatsApp Subscribers, Email Subscribers, and All Contacts.

These are built-in starting points covering the most common groupings, so you're not starting from a completely empty page.

Click + New Segment in the top-right corner. This opens the segment builder, where the real work happens.

At the top, you can rename your segment (it starts as "Untitled segment") and choose whether it's Dynamic or Static.

There's also Cancel and Save Segment buttons. Nothing is saved until you click Save.

There's an optional Description field, "Explain who belongs in this segment and how it will be used."

While optional, filling this in is genuinely useful once you have a dozen segments and need to remember why you built each one.

Right below the description, you'll see a panel labeled AI-assisted audience building (Beta).

It explains the approach clearly: build your rules first, and once they're set, AI analyzes the quality of that audience and suggests campaign opportunities based on it.

Three small indicators sit alongside this:

• Rule health — shows how many active conditions your segment currently has
• Suggestions — unlocks after your first evaluation, once there's enough data to work with
• Live estimate — refreshes automatically as you adjust your rules

If you don't want to build a rule completely from scratch, the "Quick filters — start with a preset" section gives you one-click starting points:

• Active last 30d
• High LTV
• Never purchased
• New this week
• Opened email

Clicking any of these adds a complete, ready-made rule instantly, which you can then adjust further if needed.

Under "Include contacts who match", you'll choose between MATCH ALL and MATCH ANY. This decides your segment's core logic:

• Match All means a contact has to satisfy every single condition you've set to be included
• Match Any means satisfying just one condition is enough

Conditions are organized into groups. Group 1 starts you off, and each condition inside it follows the same three-part structure:

• Property — what you're filtering on, for example, "First name," though this extends to behavioral and purchase-based properties too
• Operator — how you're comparing it, for example, "is"
• Value — what you're comparing it to

You can add more conditions to a group with + Add condition, or create a nested group entirely with + Add nested group.

Nested groups are useful for more complex logic, for example, "match all of these conditions, OR any of this separate group." Nested groups get their own AND/OR toggle, so you can mix logic types within a single segment.

Each group also shows a running summary, like "2 conditions, all conditions must match," so you can read your own logic back at a glance without having to mentally track it.

On the right side of the screen, a Live audience preview panel updates as you build:

• Matching contacts — a live count of how many contacts currently match your rules, refreshed with a timestamp
• Channel breakdown — how your matching audience splits across channels, which fills in once you have contacts to evaluate
• Top locations — location and device breakdowns, available once aggregate analytics data exists
• Sample contacts — actual example contacts matching your current rules, so you can sanity-check your logic against real people rather than just trusting the count

There's also a Query SQL section at the bottom, an expandable, more technical view of the exact underlying query your rules translate into. This is useful if you want to see precisely what's being matched under the hood.

Once your rules look right and the live preview is showing what you expect, click Save Segment.

From this point on, if it's a Dynamic segment, it'll keep evaluating automatically. Contacts will move in and out as their behavior changes, without you needing to touch it again.

If you're building your first segment and aren't sure where to begin, starting from a Quick filter, like "Opened email" or "Never purchased", is easier than building rules from a blank group.

Once you're comfortable with how conditions, groups, and Match All/Any interact, building fully custom segments from scratch becomes much more intuitive.