Audience
Turn bot identities into usable audience context.
13 identity channels · 4 consent scopes · 5 person record views · 11 insight panels. Inspect people, organize cohorts, govern retention, and move audience data with clear limits.
13
identity channels
4
consent scopes
5
person record views
11
insight panels
Audience data model
From a channel identity to usable customer context.
Audience is the operating workspace around Rylvo's organization-scoped end-user records. It keeps identity, consent, collected profiles, activity, sessions, and operator actions connected without pretending every visitor is already known.
A person reaches a bot
A channel-prefixed external ID creates or updates the organization-scoped person record.
Unknown stays separate
Opt-in web-widget collection can retain an anonymous profile without mixing it into identified people.
Context accumulates
Consent, activity, tags, bot-specific profiles, session history, and collected fields stay attached to the identity.
Operators organize
Use live segments, saved Rooms, pins, or explicit People datasets depending on whether membership should stay dynamic.
Teams take action
Inspect, correct, merge, export, govern retention, or pass eligible audiences into Broadcasts.
People workspace
Find the person, then see the whole operating history.
Browse the directory efficiently, open a full record, keep anonymous collection separate, and resolve duplicate known identities with an audited merge.
AUDIENCE / PEOPLE
PRODUCTION · TEST IDENTITIES HIDDENFirst seen
12 Mar 2026
Last seen
2 hours ago
Consent
Full · 730-day
Tags
vip, at-risk
Collected profile · support-bot
order_id
NS-40912
issue_type
refund
•••••• redacted
Browse production people with server-driven pages
The People workspace reads the canonical end-user directory and keeps test identities out of the production view by default.
Free-text and forgotten-state matching can be applied over the loaded page, while supported structured filters are pushed into the server query.
Organize and collaborate
Three ways to keep an audience, based on how membership should change.
Segments, Rooms, and People datasets are related but deliberately different. Choose dynamic resolution, a shared working view, or explicit durable membership.
Segments
Saved audience predicates resolve against current data and maintain a materialized member index.
- Live preview before save
- Activity, channel, conversation, checkpoint, field, and completeness predicates
- Reusable for eligible broadcasts and exports
Audience Rooms
Store a People-page filter state for an investigation, cohort, account list, or team-owned operating queue.
- Named filters and assignees
- Pin people with notes
- Room and per-person comment threads
- Archive without immediately discarding discussion history
People datasets
Create a durable named collection of people from the People page or from every pin in an Audience Room.
- Deduplicated membership by external ID
- Separate from trace datasets by kind
- Archive, update, or remove individual members
- Uses the workspace dataset allowance
Segments can seed Broadcasts when the plan supports it
Saved membership alone does not override consent, channel delivery rules, campaign limits, or the Broadcasts plan gate.
Consent, privacy, and retention
Controls that support a privacy program—not a compliance promise.
Make consent state visible, restrict sensitive actions, retain data for the intended window, and preserve an audit trail around operator changes.
Four per-person consent scopes
The consent scope controls personalization behavior and its per-user retention default.
None
Per-conversation context only
0-day personalization retention
Short-term
Temporary cross-session memory
Up to 7 days
Personalize
Long-term user model without raw-turn indexing
365-day per-user default
Full
Personalization plus episodic recall
730-day per-user default
Forget and delete
Authorized actions cascade through the selected Rylvo records, update audience counts, invalidate sessions where applicable, and record the operator action.
Plan-aware retention
Use the plan window, shorten it for stricter policy, or extend eligible retention with an add-on. A scheduled sweep purges expired data.
Consent and action evidence
Per-person audit history and the consent workspace show decisions and operator actions without claiming they replace legal process.
Lifecycle webhook
Send consent changes, removal events, imports, and session resets to one HTTPS destination, with HMAC signing when a secret is configured.
Audience insights
See where people engage, consent, and finish the profile.
Audience Insights turns current person and profile data into an organization-wide or bot-scoped snapshot. Cached reads keep the page responsive; authorized users can request a fresh computation.
Snapshot, then recompute when the decision needs fresh numbers
Mutations attempt to invalidate cached snapshots. The Insights page also shows snapshot age and exposes a recompute control.
Eleven views
The bot leaderboard is organization-wide. Other analytics can be scoped to the currently selected bot where the underlying data supports it.
Import and export
Bring the directory with you. Take the useful slice back out.
Use CSV for audience migration and segment portability, with a preview before writes and explicit limits instead of an unbounded background promise.
Four-stage CSV import
Upload a file up to 15 MB, map columns, preview create/update/error counts, then commit. The server also enforces a plan-specific row limit.
External-ID idempotency
The same external ID updates an existing record rather than creating a duplicate. Updates merge display name and tags without replacing consent or rolling statistics.
Bounded exports
Export a saved segment or the current audience filter to CSV, up to 5,000 resolved rows per export. A single person's complete record can be exported as JSON.
Profiles & Schema shows the collection contract across bots
Review the flattened field catalog, per-bot field types, required flags, checkpoints, and validator state. Seeing repeated keys side by side helps operators spot naming drift; Rylvo does not claim automatic synonym detection.
Access and plans
Let more people see the audience than change or erase it.
Audience uses separate capabilities for viewing, field correction, editing, session management, exports, and destructive privacy actions. Capacity and retention then vary by plan.
Viewer
Browse Audience pages, people, anonymous captures, segments, insights, consent state, and schema information.
Operator
Correct an individual collected profile field through the scoped override capability.
Admin and owner
Manage tags, segments, Rooms, imports, sessions, exports, consent, retention settings, and forget/delete workflows.
Audience is included across plans with different allowances
Free through Enterprise differ in saved Segment limits, Room limits, import rows per job, and included audience retention. Broadcasts requires Pro or higher.
FAQ
Identity and privacy boundaries, plainly stated
The distinctions that matter before a team starts using Audience as an operating workspace.
Does Audience automatically identify the same person across channels?
No. Identities are keyed by channel-prefixed external IDs. An authorized operator can explicitly merge two known records; the primary identity wins conflicts and forgotten records cannot be merged.
Are anonymous widget visitors mixed into the People directory?
No. When anonymous collection is enabled, those bot profiles use synthetic subjects and appear in Anonymous Captures. Production People excludes internal test identities by default as well.
What is the difference between a Segment, Room, and People dataset?
A Segment is a dynamic query, a Room is a saved operational view with assignees, pins, and discussion, and a People dataset is an explicit named membership list. They solve different continuity needs.
Does Audience guarantee GDPR compliance?
No product can guarantee compliance by itself. Audience provides consent scopes, access controls, audit history, export, forget/delete workflows, configurable retention, and scheduled purging that can support your privacy program.
What happens when I re-import the same CSV?
The external ID is the document key, so a repeated import updates the existing person instead of creating a duplicate. Existing consent and rolling activity statistics are not overwritten by the update path.
Can every plan use Broadcasts from Audience?
No. Audience management is available across plans with different capacity and retention limits, while Broadcasts is a Pro-or-higher capability. Consent, channel, and campaign rules still apply at send time.
Give every person record a clear operating context.
Browse identities, inspect sessions and profiles, organize dynamic or explicit cohorts, govern consent and retention, and move data in or out with clear limits.