Segment filter reference

Segments are built from rules. Each rule pairs a field with an operator and a value.

Device & location

• Device — desktop, mobile, tablet.
• Browser — Chrome, Firefox, Safari, Edge, Opera, Samsung Internet.
• OS — Windows, macOS, iOS, Android, Linux, Chrome OS.
• Country — ISO code or name.

Behaviour

• URL path — pages visited (e.g. /checkout).
• Session count — number of sessions per visitor (numeric comparisons).
• User ID — your identified user identifier.

Acquisition (first-touch)

• UTM source / medium / campaign / term / content.
• Traffic channel — Direct, Organic search, Paid, Social, Email, Referral.
• Source host — the referring domain (e.g. google.com).

Events and their properties

Anything your site sends with Spectry.track() is filterable, including events tracked only from code:

• Event name — the name you passed to track(), e.g. purchase.
• Event properties — the values you sent alongside it, listed under Event properties in the field picker.

Visitor context

Attributes set with Spectry.setContext() — plan, account type, locale and anything else you attach to the visitor — appear under Visitor context. Context is stamped onto each event as it is captured, so a rule reads the value as it was at the time of the event, not as it is today.

The event-property and context lists are built from the last 7 days of your traffic, so a property nobody has sent recently is not offered. Rules already saved for it keep working.

Operators

• equals / not equals — exact match.
• contains / not contains — substring match.
• is one of — match any value in a list.
• >, ≥, <, ≤ — numeric comparisons (for numeric fields like session count, and for numeric event properties like order_total).
• is set / is not set — whether an event property or context attribute was present at all. These take no value, and are the only way to tell "absent" apart from an empty string or 0.

Rules about events are matched against the same event, not merely the same visitor. Event name is purchase and order_total > 100 together mean one purchase worth over 100 — not a visitor who bought something and separately triggered a large order.

Where a segment can be applied, you'll find a segment dropdown (showing each segment's size) alongside an "All visitors (no segment)" option.