Prevenue Docs

Event Mapping and Signal Readiness

Confirm event semantics so Prevenue can evaluate the right signal families.

Event mappings tell Prevenue what each source event means. Signal readiness uses those mappings, account identity, event history, and requested revenue motions to decide whether a signal family can be evaluated.

Open Events to review mapping suggestions in the event table. Click a row when you need payload context, account facts, sentiment controls, or advanced mapping fields in the event drawer.

Mapping basics

A mapping connects category and sentiment defaults for an exact workspace + source + raw event name pair:

  • Source, such as direct, stripe, posthog, segment, or web_sdk.
  • Raw event name, such as upgrade_clicked.
  • Semantic type, such as upgrade_intent.
  • Category status, such as confirmed, needs_review, or ignored.
  • Sentiment default, such as positive, neutral, negative, unknown, or ignored.
  • Sentiment score, confidence, source, and review status.

Confirmed category mappings can improve readiness. Ignored category mappings keep noisy, diagnostic, or irrelevant events from influencing signals.

Sentiment is reviewed independently from category. ignored sentiment means Prevenue does not apply the mapped sentiment fallback; events without explicit sentiment remain unknown.

Semantic types

Use the semantic type that describes the customer meaning of the event:

Semantic typeUse it for
usageValue-producing product usage.
activationFirst-value or setup-completion milestones.
limit_frictionQuota, paywall, or usage-pressure blockers.
upgrade_intentPricing, premium feature, or upgrade interest.
downgrade_intentDowngrade flow or contraction interest.
cancel_intentCancellation or account deletion intent.
checkoutCheckout starts, completions, or abandonment.
topupAdd-on, credit, or usage top-up behavior.
billing_lifecycleSubscription, invoice, payment, billing-status, or plan lifecycle changes.
team_expansionInvites, seats, users, roles, or team growth.
loginLogin and engagement cadence.
support_frictionSupport, setup, sync, or product friction.
integration_connectedIntegration, OAuth, or destination setup completion.

When a source event does not represent a meaningful customer action, ignore it.

Sentiment mapping

Mapped sentiment fills gaps when an event payload omits sentiment or sends unknown. Explicit event payload sentiment wins over the mapping by default.

Precedence is:

  1. Explicit event payload sentiment.
  2. Confirmed or auto-applied mapping sentiment when the mapping is not ignored and the mapped sentiment is not unknown.
  3. unknown.

Use unknown when sentiment is unavailable. Use neutral only when the product signal is explicitly neutral. Mapping propagation preserves existing explicit event sentiment unless a user intentionally chooses overwrite in the Events drawer.

AI mapping analysis

Prevenue can suggest category and sentiment mappings based on event names, payload shape, known product patterns, semantic type, event kind, and redacted property keys. Treat suggestions as reviewable guidance.

High-confidence sentiment suggestions can be auto-filled for review. Lower-confidence sentiment suggestions remain unknown or ignored until a user changes them.

Before accepting a suggestion:

  • Check the event name and source.
  • Inspect recent payload examples.
  • Confirm the event affects the right revenue motion.
  • Confirm the event is customer evidence, not a diagnostic.
  • Change the semantic type if the suggestion is close but wrong.
  • Check whether sentiment is explicit, inferred, AI-suggested, integration-provided, or unknown.

Signal readiness states

StateMeaning
BlockedRequired setup is missing.
PartialSome useful evidence exists, but more setup or evidence is needed.
WaitingSetup exists, but the workspace needs more recent qualifying events.
Ready to previewThe signal family can produce reviewable examples.
ActiveThe signal family is ready for normal operation.
Not requestedThe workspace did not ask for this motion or signal family.

Readiness is about setup state. It is separate from prediction readiness, which describes data/model sufficiency caveats.

Review workflow

  1. Open Events.
  2. Review suggestions marked needs review.
  3. Confirm high-value category mappings.
  4. Assign or ignore mapped sentiment defaults.
  5. Use overwrite explicit sentiment only when intentionally rewriting existing event sentiment.
  6. Ignore diagnostic and low-value events.
  7. Run mapping analysis again if source data changed.
  8. Re-run the Revenue Signal Scan.
  9. Open Signals and inspect detail pages before routing.

When to use mapped events

Use confirmed mappings as Funnel targeting inputs and intelligence evidence after identity and readiness are healthy. A deterministic Funnel may also use an explicit typed event contract without requiring every event family to be categorized first.

If readiness is blocked, fix source events, identity, and mappings before adding learned thresholds. Default safe mappings when confidence is strong and reserve manual review for uncertainty or material exceptions.