Data Intelligence Framework

Data Intelligence Framework (a customer-data preparation and decision model) connects identity, consent, operational status and timestamped events into usable lifecycle context. It defines how source data becomes a reviewed decision input, so planners and channel integrations can work from consistent meanings instead of disconnected exports.

Updated

The problem it removes

A source may call a profile active because the account exists, while an ESP uses the same word for a contact that can receive email. A second tool may treat a recorded open as engagement, even though the team's measurement policy requires a click, reply or purchase. Copying those fields without their definitions creates unreliable segments.

Identity and permission gaps make the problem worse. The same address can appear under several properties while its unsubscribe remains in only one. The framework connects these meanings before automation uses them. It helps the team identify which source is authoritative and which missing or conflicting input should prevent the proposed action.

How it works

  1. Map the sources and define the fields used by the first application. Identify the profile key, subscription scope, event timestamps and evidence that establishes permission for the intended brand and channel.

  2. Connect identity and subscriptions through Unified Profile Graph. Preserve source context so operators can understand both the match and the permission boundary.

  3. Structure operational fields through Live Marketing Status and behavior through the Event-Centric Timeline. Distinguish current eligibility from historical interest and model predictions.

  4. Give Intelligent Planners and the Decision API the agreed inputs. Validate a complete decision path, including the exclusions, before applying the same mapping to a larger audience.

Unified customer profile graph One reusable profile connects multiple sites, subscriptions, consent records, events, live status, customer segments, and channel preferences. UNIFIED PROFILE GRAPH One person. Multiple sites. One reusable profile. CUSTOMER 0428 Unified Profile Consent current · status safe · 4 identities linked AI-ready Email IdentitySite A Subscription Site B SubscriptionConsent Source EventsLive Status Segment / ClusterChannel Preferences ID resolvedWeekly Digest Product UpdatesForm · 2026-07-18 19 time-aware actionsEligible · interested High-intent readerEmail · mornings Multi-siteConsent-awareReusableAI-ready

The existing model includes status.hard_bounce, status.unsubscribed, status.series_interest, status.cluster_id and status.timezone_validated. These fields answer different questions. A useful implementation review documents their sources and checks whether an update reaches the next decision, rather than accepting a populated column as proof that the workflow is correct.

Review data minimization and brand boundaries as part of the mapping. The platform states US data residency; access and the data needed for the selected application should be scoped before integration. A shared data model does not imply shared permission across every property. The decision must still use the subscription and suppression state that applies to its intended action.

Keep the input definitions available to the people reviewing each decision.

Where it shows up

Lifecycle Decisioning combines Data Intelligence Framework, Intelligent Planners and Decision API in the full waterfall. Multi-Brand Profile Unification demonstrates why identity and distinct permissions need to remain connected. Implementation Services maps the sources and verifies the first application; the platform hub explains how all eight components fit together.

What it is not

The framework is not a replacement for a data warehouse, CDP or ESP, and it does not declare competing tools incapable of managing customer data. Its purpose is to make the required inputs consistent enough for a reviewed decision. It cannot recover missing consent evidence or establish incremental business performance merely by cleaning records. Those questions need their own evidence and measurement plan.

FAQ

Where should we start with inconsistent fields?

Start with the fields required by one application. Define their meaning, source and permission scope, then inspect representative eligible and suppressed profiles before broadening the mapping.

Does identity matching establish permission?

No. Identity and permission are different inputs. A connected profile must retain the source and scope of each subscription so a decision can evaluate the intended brand and channel.

Can the framework work with Klaviyo or beehiiv?

The guides explain platform-specific contact states and operating choices. The actual integration is scoped around available interfaces and your requirements, with current provider behavior checked before activation.

What demonstrates a successful implementation?

A traceable flow from source data to the correct action or suppression for a defined cohort. Clean columns and a working import are useful checks, but the decision behavior must also be observed.

Read more

Read the Klaviyo inactive-profile guide and the beehiiv inactive-subscriber guide for examples of why platform states, billing and permission decisions need explicit definitions.