A customer data platform (CDP) is software that collects customer data from many sources, such as websites, apps, CRM, billing, support and marketing tools, matches records that belong to the same person or account, and builds unified customer profiles. It then makes those profiles available to other systems so marketing, sales and service teams can act on a consistent view of each customer.
At a glance
- A CDP pulls data from many systems, links it into profiles and shares those profiles with other tools.
- It usually handles both known customers (with an email or account) and, depending on the product, anonymous website or app visitors who later identify themselves.
- It is often owned by marketing but increasingly feeds customer service and analytics.
- Some CDPs store their own copy of data; others work on top of an existing data warehouse.
- Because it centralizes personal data, consent, retention and access rules matter; obligations vary by jurisdiction.
What problem it solves
Customer information is usually spread across many systems that don’t agree. The website knows what someone browsed, the CRM knows their account manager, billing knows what they pay, the support platform knows their open cases, and the email tool knows which campaigns they opened. Each team sees a fragment, and joining the pieces typically needs IT or data engineering for each request.
A CDP joins them. It gives marketing and service teams a single profile for each customer and lets them build audiences (“customers who renewed last year and opened a support case this month”) without a custom data project each time. It can also send those profiles back into the tools where work happens, so a contact center agent sees recent web activity, or an email campaign can exclude customers who already bought.
How it works
Collect. Connectors and tracking code bring in data from websites, apps, CRM, e-commerce, billing, support platforms, marketing tools and offline sources such as point-of-sale.
Unify. Identity resolution links records that belong to the same customer, using identifiers such as email addresses, phone numbers, account IDs or device IDs. This step decides how accurate the profiles are and is often where implementations struggle.
Enrich and segment. The CDP calculates attributes (lifetime value, last purchase, propensity scores, depending on the product) and lets users build segments.
Activate. Profiles and segments are sent to other systems: advertising platforms, email tools, the CRM, the contact center platform or analytics. This can be in batches or close to real time, depending on the product and destination.
Govern. A CDP can attach consent and preference attributes to profiles, for example to mark someone who opted out of marketing. Whether that person is actually suppressed depends on the CDP matching all of their records, each destination integration passing the attribute along, how it is configured, and the downstream tool honoring it. Buyers should verify propagation and enforcement in every destination and reconcile or audit regularly; the company remains responsible for honoring opt-outs, not the software. Deletion requests are a separate job: they need to be carried out and checked in the CDP itself, including profiles, derived attributes and retained copies, in the source systems that feed it, and in the destinations it has sent data to, since deleting downstream does not remove source records. Sound data governance covers who can see what, how long data is kept and how deletion requests are handled across all of those systems. Requirements depend on the privacy laws that apply to your customers, such as the GDPR in Europe or state privacy laws in the US.
When it matters for buyers
- When marketing, service and sales each have a partial view of customers and want to work from one.
- When building customer journey maps and you need data to show where customers actually go and drop off.
- When consolidating marketing tools, since a CDP can replace several point integrations.
- When privacy requests and consent are hard to manage across many systems.
- When you already have a data warehouse, which changes whether you need a full CDP or a lighter tool that works on top of it.
For tools that sit around customer data, see our customer relationship management and analytics and business intelligence pages.
Questions to ask vendors
- Which of our systems have ready-made connectors, and which need custom work?
- How does identity resolution work, and can we see and adjust the matching rules?
- Do you store a separate copy of our data, or can you run on our existing warehouse?
- How quickly can profile updates reach destination systems such as the contact center or email tool?
- How are consent and preferences recorded and passed to each destination, and how can we verify and audit that each one honors them?
- How is a deletion request carried out across inbound sources, the CDP itself (profiles, derived data, backups and retained copies) and outbound destinations, and what evidence do we get that it is complete?
- Where is data stored and processed, and what options do you have for data residency?
- How is pricing calculated: by profiles, events, data volume or destinations?
How it differs from a CRM
A CRM is a working system for sales and service staff: accounts, contacts, deals, cases and activity records, much of it entered by people. A CDP is a data system: it gathers data automatically from many sources, including behavior that often never appears in the CRM, builds profiles and passes them to other tools. A CRM typically answers “what is happening with this account?” for the person handling it; a CDP answers “what do we know about this customer across every system?” and shares that answer with other tools. The two are often connected, with the CDP feeding profile data into the CRM and drawing CRM records back out.