Skip to main content

Customer Data Platform

The customer data platform layer in DriveMetaData is the foundation for reporting, AI insight, segmentation, and activation. It brings customer signals from multiple systems into a unified, usable customer view.

What the CDP layer solves

Customers rarely interact through one clean channel. A single journey can include an ad click, anonymous web session, mobile app install, email open, purchase, support interaction, and offline conversion. Without a shared customer view, teams see fragments instead of behavior.

The CDP layer helps teams:

  • Connect behavior, campaign, CRM, product, and transaction data.
  • Resolve identities across permitted identifiers.
  • Build profiles that update as new events arrive.
  • Make reporting and activation use the same customer definitions.
  • Reduce manual reconciliation across tools.

Customer profile structure

A typical profile includes:

Profile areaExamples
IdentifiersUser ID, anonymous ID, CRM ID, device ID, email hash
TraitsRegion, lifecycle stage, account type, language, preferences
EventsPage view, app open, signup, add to cart, purchase, renewal
Campaign historySource, medium, campaign, ad group, creative, click metadata
PredictionsConversion likelihood, churn risk, LTV band, engagement momentum
Activation statusAudience memberships, channel eligibility, suppression flags

Identity resolution

Identity resolution connects related signals into one profile when the data supports it. For example, an anonymous web visitor can become a known profile after signup, or a mobile app event can be connected to a CRM user after login.

Good identity rules should be:

  • Deterministic where possible.
  • Reviewed by analytics and privacy stakeholders.
  • Documented for reporting users.
  • Tested with edge cases such as shared devices, duplicate emails, and account transfers.
warning

Do not merge identities only because two records look similar. Use approved identifiers and consent-aware rules defined by your organization.

Real-time and batch data

DriveMetaData-style workflows often combine both:

  • Real-time events for journeys, segment updates, fraud signals, and behavior-triggered actions.
  • Batch data for CRM fields, revenue adjustments, offline conversions, refunds, and warehouse enrichment.

When a report seems delayed, confirm whether the underlying source updates in real time, on a schedule, or after an internal approval workflow.

Data quality checklist

Before using a dataset for reporting or activation, check:

  • Event names are consistent across web and app.
  • Required event properties are present.
  • Timestamps are in the expected timezone.
  • Revenue and currency fields are normalized.
  • Test traffic is labeled or excluded.
  • Consent and suppression fields are available to activation workflows.
  • Duplicate records are handled before KPIs are published.

How the CDP layer supports other features

  • Attribution uses profile and campaign history to connect touchpoints to outcomes.
  • AI analytics uses behavioral patterns to identify intent, churn risk, and likely value.
  • Segmentation uses profile traits, events, predictions, and eligibility rules.
  • Fraud protection uses traffic, device, click, install, and event patterns to flag suspicious activity.