TEALIUM · SEGMENT · RUDDERSTACK
CDP Implementation
A customer data platform that actually ships, not one that lives in a roadmap deck.
A CDP implementation is the program that lands a single customer profile across web, app, and CRM, with identity resolution, consent governance, and activation built into the same system. End-to-end programs delivered on Tealium, Segment, and RudderStack, from data model design through audience activation.
Outcomes
One profile, every surface
Web, app, CRM, and product behaviour resolved into a single customer view your teams can trust.
Identity that holds up
Deterministic and probabilistic matching, decay rules, and an audit trail you can defend in a privacy review.
Consent baked in
TCF and GPP signals propagated to every destination. No quiet leaks, no after-the-fact retrofits.
Activation tied to revenue
Audiences flowing into ads, email, push, and product experiences from day one of activation, not after the platform is fully finished.
How the engagement runs
Discovery
Audit your data sources, identity logic, consent posture, and existing reporting. Map the current state on one page so the rest of the program has a shared reference.
Data model and schema
Define the events, traits, and audiences that matter. Draft a naming convention and a contract that survives the next vendor change.
Identity and consent
Resolution rules, household and account stitching where useful, and a consent state model aligned to TCF v2.2 and GPP. Documented for the DPO.
Build and integrate
Source SDKs, server-side connectors, and destination integrations live in production behind a release plan with rollback.
Activate and handover
First audiences shipped, dashboards in place, and the internal team trained against a runbook. The program does not collapse when I leave.
What you get
- Documented event taxonomy and trait dictionary
- Identity resolution graph with confidence rules and decay
- Consent state model mapped to TCF v2.2 and GPP
- Source and destination integrations live in production
- Activated audiences with a measurement plan attached
- Internal runbook covering ops, QA, and incident response
- Enablement sessions for marketing, data, and engineering
Where this work has shipped
Selected programs. Names listed with permission.
Vodafone Italy
CDP Strategy & Implementation
Unipol
Real-Time CDP Data Modeling
De'Longhi
Data Strategy & MarTech Advisory
ho.mobile
Consent & Compliance Orchestration
Common questions
Which CDP should I choose?
Tealium, Segment, and RudderStack each fit different shapes of company. Tealium leads on enterprise compliance and EU data residency. Segment suits product-led SaaS teams. RudderStack appeals to teams who want a warehouse-native architecture. The right answer comes from your stack and team, not a vendor leaderboard.
Do I actually need a CDP, or is a tag manager enough?
GTM solves tag deployment. A CDP solves identity, audiences, and activation. If your problem is fragmented user profiles across web, app, and CRM, a CDP is the right tool. If your problem is just shipping pixels reliably, a tag manager is enough and a CDP is overkill.
Can you migrate us off our current CDP?
Yes. Migrations work best when paired with a data model revision rather than a like-for-like swap, since the previous model usually carries a few years of compromise. Past replatforming work covers Tealium, Segment, and mParticle, including parallel-run cutovers.
How do you handle GDPR and consent?
Consent runs through the CDP as a first-class signal, not a sidecar. Each source emits a consent state, and destinations only fire when permitted. The implementation aligns to TCF v2.2 and GPP, with audit trails your DPO can hand to a regulator without rewriting them.
Will you work with our existing engineering team?
Yes, almost always. The handover plan assumes your team owns the platform at the end of the engagement. The work includes documentation, paired build sessions, and enablement so the program does not collapse on rollover.
What drives scope and cost on a CDP program?
Scope is driven by the number of source systems, the destination footprint, identity complexity, consent and regional governance, and the depth of activation required at launch. Each of those moves the engagement substantially. Cost and timeline are scoped during discovery against the specific shape of your stack, never quoted blind.
What is the difference between a real-time CDP and a warehouse-native one?
A real-time CDP, like Tealium AudienceStream, resolves identity and activates audiences inside the platform itself, with low-latency triggers. A warehouse-native CDP, like RudderStack or Hightouch-style architectures, leaves the data in your warehouse and syncs audiences out via reverse ETL. Real-time wins on activation latency and operator UX. Warehouse-native wins on data ownership and modelling depth.
Want to see if this fits?
A short intro call to talk through scope, the questions you are carrying, and whether the work is a sensible match.