Saved Segments

Saved Segments let you build a named filter once and re-apply it across Globe, Pages, Sources, Journeys, Users, and Funnels in a single click. They go beyond regular filters in three ways:

1. They're persistent. A regular filter lives only in the URL and disappears when you close the tab. A segment lives in your team's workspace.
2. They support OR logic. Regular filters AND every condition together. Segments support up to 3 AND-groups joined by OR — e.g. (paying users from Germany) OR (visitors who scrolled past 75% on /pricing).
3. They unlock cohort dimensions that regular filters don't expose: user type (anonymous / identified / paying), scroll depth, sessions count, page count, conversion status, custom traits, and the AI visitor score.

Building a segment

Open the Filter funnel icon in the dashboard header. At the bottom of that dropdown, the Saved Segments section lists every segment for this website. Click any row to load it instantly. Click + New segment to open the builder.

The builder is a vertical list of groups. Within each group, conditions are joined with AND. Between groups, conditions are joined with OR.

[ Group 1: country is "Germany" AND user_type is "paying" ]
                          OR
[ Group 2: scroll_depth_max ≥ 75   AND page is "/pricing" ]

V1 caps:

Once saved, the segment appears in the dropdown and is auto-applied. A blue chip in the dashboard header shows the active segment name; click the (×) to clear.

Managing existing segments

Hover any row in the Saved Segments list:

• Pencil ✏️ opens the segment in the builder, pre-filled, so you can adjust conditions or rename it without recreating from scratch.
• Trash 🗑 deletes immediately. A green confirmation appears at the bottom of the screen.

Clicking a saved segment loads it across every chart and panel on the dashboard. The active period selector (e.g. Last 7 days) still applies — the segment defines who, the period selector defines when.

Why the value counts in the builder match what you see

Each option in the dimension dropdowns shows a count (e.g. United States · 230). That count reflects the dashboard's current period, not a fixed 30-day window. If you're viewing Last 7 days, the count is the 7-day count. Applying the segment will produce numbers that line up with what you saw next to each option.

A common pitfall — AND inside one group ≠ OR

If you add country is Italy and country is Australia to the same group, you've built country IS Italy AND country IS Australia — which matches zero visitors, because no visitor is from both countries simultaneously. To match either, put each country in its own OR group. The builder shows an amber heads-up when it detects this pattern.

Dimensions

Saved segments expose every dimension available to regular filters, plus 11 new ones:

URL & Source (existing)

hostname, page, entry_page, country, region, city, channel, referrer, campaign

UTM (new in P1-1)

utm_source, utm_medium, utm_content, utm_term

System (existing)

browser, os, device

Engagement (new in P1-1)

Identity (new in P1-1)

A visitor is paying if they have any payment-provider customer ID (Stripe / Lemon Squeezy / Polar / PayPal), total_revenue > 0, or an active subscription_status. A visitor is identified if they have an email or customer_id sent via zenovay.identify(...). Otherwise they are anonymous.

Custom traits (new in P1-1)

zenovay('identify', {
customer_id: 'cust_42',
email: '[email protected]',
// Custom traits — top-level keys are queryable via custom_prop:<key>
custom_attributes: {
  plan: 'enterprise',
  industry: 'fintech',
  seats: 25,
},
});

In the segment builder, the dimension dropdown will offer Trait: plan, Trait: industry, Trait: seats under the CUSTOM TRAITS section.

Operators

Examples

Paying users from Germany who bounced on /pricing

Three conditions, one group:

Group 1 (AND):
  user_type   is        paying
  country     is        DE
  page        is        /pricing

Engaged users OR converters

Two groups, joined by OR:

Group 1 (AND):
  scroll_depth_max  gte  75
  sessions_count    gte  3

OR

Group 2 (AND):
  has_converted     is_true

Enterprise plan trait + Chrome on macOS

Group 1 (AND):
  custom_prop:plan  is        enterprise
  browser           is        Chrome
  os                is        macOS

Public dashboards

Saved segments are never applied via share-token authentication. The public view honours regular URL filters but ignores ?segment_id= and ?segment_def= query parameters by design — the audience of a share link should not see your team's named cohorts.

If you need to share a single composed view publicly, the cleanest approach is:

1. Apply the segment privately.
2. Recreate the same conditions as flat filters in a public-friendly URL.
3. Share the resulting URL.

Permissions

Viewers can apply existing segments but cannot create or modify them.

Performance & caps

• Per-website segment count is capped at 50 to prevent runaway list growth. The same cap applies on every plan; segments are not paywalled.
• Resolution model: when a segment loads, each AND-group is resolved server-side to a list of matching visitor IDs, then the lists are unioned across groups. Result limit per group is 50,000 visitor IDs — adequate for nearly every use case.
• Aggregate dimensions (sessions_count, pages_count) execute via SECURITY DEFINER Postgres helpers (segment_visitors_by_session_count, segment_visitors_by_page_count). Page-count lookups consider the last 90 days.
• Custom traits are looked up via JSONB key access on identified_users.custom_attributes. Top-level keys only — nested paths are not supported in V1.

Related

• Dashboard Overview — how filters and segments compose with the dashboard's date range.
• Visitor Identification — how to send zenovay.identify(...) so user_type and custom traits are populated.
• Visitor Value Scoring — how ai_score is computed.