Every Connection
Has a Purpose

Define consistent events and properties so browsing, purchases, and product activity can inform relevant messaging decisions.

Connect each recorded action with useful context, including timing and product details, while respecting collection preferences.

Review event naming, identifiers, and delivery quality before allowing customer activity to trigger downstream campaigns.

Connect known identifiers across supported touchpoints to build customer context without assuming every anonymous interaction belongs together.

Use relevant profile attributes and observed activity to describe audiences that support a clear communication purpose.

Review identity rules and destination requirements before activating audiences, keeping collection preferences separate from messaging permissions.

Agree what a predicted outcome means before evaluating whether available historical data can support a useful model.

Group eligible customers by estimated churn risk to explore differentiated retention messaging and compare outcomes against controls.

Review prediction quality and audience performance regularly, treating model outputs as estimates rather than confirmed customer intentions.

Use supported customer attributes and event properties to adapt message content to the context available at send time.

Define clear content conditions so different audiences receive appropriate information without relying on unsupported assumptions about individual intent.

Provide useful fallback content when attributes are missing, and preview representative records before approving personalized messages.

Define audience eligibility and entry triggers so each journey starts from an explicit customer action or business context.

Branch journeys using observed actions and evaluation windows, allowing customer behavior to influence which message comes next.

Use deliberate delays, reentry rules, and communication limits to coordinate contact without overwhelming people across competing journeys.

Guide new customers toward meaningful first actions, adapting followup when they have already completed the intended milestone.

Use observed feature activity to identify appropriate education opportunities and avoid repeating guidance customers no longer need.

Define inactivity criteria and exit conditions before testing reengagement messages, with suppression rules that respect current communication preferences.

Choose meaningful engagement events and assign documented weights so teams can understand why a customer receives a score.

Reduce the influence of older activity using supported decay rules, keeping current engagement distinct from historical interaction volume.

Review score distributions and outcomes before setting operational thresholds, then monitor whether prioritization remains useful as behavior changes.

Align recorded marketing interactions with supported contact and deal associations so revenue reporting has a traceable data foundation.

Compare supported attribution models to understand how different credit rules change the reported contribution of recorded touchpoints.

Document tracking coverage, conversion definitions, and missing interactions before using attributed revenue to inform future marketing investment decisions.