Knowledge · IT · Data · DX

CDP · Customer Data Platform

Customer Data Platform

📂IT · Data · DX⏱2-min read📊Practitioner🔗4 related

What is a CDP?

In one line
A marketing Customer Data Platform (Customer Data Platform) that unifies customer data scattered across channels on a real-name basis to build persistent customer profiles.
⚡ Summary in 3 lines
  • Unifies scattered customer data on a real-name basis
  • Persistent customer profiles
  • The foundation of personalization and omnichannel

Turning on a CDP isn't about adding one more system; it changes how a company counts 'how many customers we have.' Until then, a web cookie visitor, an app device ID, an order-system purchaser, and a call-center inquirer were four different people. After identity resolution, the four merge into one, and the denominator of metrics like repurchase rate, average order value, and churn rate changes on the spot. It's normal for the member count to drop right after adoption. Customers haven't vanished; the duplicates have been removed.

The biggest source of confusion in practice is the boundary with CRM, DMP, and MA (marketing automation). A CRM is a system of record for contracts and transaction history, a DMP groups anonymous cookies into segments for advertising, and MA executes campaign scenarios. A CDP sits between them as a repository that attaches behavioral data and real-name data to a single profile. How you attach them also varies. Deterministic matching links records by a reliable key such as a member ID or hashed phone number; probabilistic matching infers links from device, IP, and behavior patterns. If the data will be used for customer support and complaint handling rather than ad targeting, it's safer not to mix probabilistic match results into the profile. These days a composable approach, layering only the functions on top of a data warehouse, is also an option.

The biggest misstep is 'buying the tool first.' If you pour in data while each channel has a different ID scheme and contact details are formatted inconsistently, you get messy data merged into messy profiles. If the match rate stays in the 40s, any segment you pull is only half-complete. Consent scope is another trap. Combining collection and usage consents obtained separately per channel, without checking them, risks violating the Personal Information Protection Act, and projects have stalled for months on a legal ruling that 'this data cannot be used.' Finally, if the CDP is kept as a marketing-only tool, sales and customer service keep looking at the old screens, and customers get inconsistent service from the same company.

Courses that cover this term · AI & DX
Recommended courses · Recommended instructors · Quick quote
🧮 Mini quote · Based on instructor fees
About KRW 1.2 million (≈ US$900)2-hour talk · up to 20 people · Korean
Get a quote for these requirements →
💡 In plain English
Stitching a customer's many traces into one person, so fragments start to look like 'one customer.'
Why it matters in practice
A foundation for turning scattered customer data into a single asset. The core of data marketing in the cookieless era.
Origin and sources

Marketing analyst David Raab established the concept in 2013 by coining the category 'Customer Data Platform,' and it spread widely as third-party cookies were phased out.

Source · David Raab, who coined the term CDP (2013) · View original
An example

Consider a hypothetical case: Company A, a beauty retailer with 520 employees and KRW 240 billion (≈ US$179,560,100) in annual revenue. It has 1.9 million online store members and 46 stores nationwide, but the online store, app, and in-store membership programs lived in separate systems, so the same person showed up under three IDs. The CDP project began when the repeat-customer figure pulled by the CRM team was off by 12% from reported revenue.

  1. The data team lead spent the first 3 weeks on cleanup, not integration. After finding five different phone number formats and store memberships holding only the last digits, they set a rule: 'unify the keys before joining anything.'
  2. They made deterministic matching the rule, linking records only through member ID and hashed phone number. The initial match rate was 51%; after making full phone number entry mandatory on the in-store tablet sign-up screen, it rose to 83% within 4 months.
  3. The legal team and the privacy officer compared consent wording across channels and excluded 270,000 in-store sign-ups from before 2019 from mailings. They applied a dual flag: the records go into the profile, but marketing use is blocked.
  4. The head of customer service set a condition from the other side. 'If the agent screen doesn't show the last three orders, the CDP is pointless.' Once a unified-profile widget was added to the support console, average handling time dropped from 4 minutes 10 seconds to 2 minutes 50 seconds.
  5. Six months later, deduplication cut the member count from 1.9 million to 1.58 million, but 220,000 customers who buy both online and in stores became visible for the first time. A campaign aimed only at them converted at 2.4x the previous rate, and quarterly repeat-purchase revenue grew 9%.
Related terms
Training linked to this term
📚Digital TransformationNanum Consulting corporate training · In-person / OnlineView Course →
Need in-house training on this topic?
Get a curriculum and quote tailored to your needs right away.
Ask About Training
Last updated 13 Sep 2026 · Reviewed by John Kim

FAQ

A CDP is for long-term customer management using identified, first-party data; a DMP (Data Management Platform) is for short-term advertising using anonymous, cookie-based data.
A CRM mainly manages sales relationships; a CDP focuses on unifying data from every touchpoint for marketing personalization.
A CDP is closer to 'identity-based unification of first-party customer data,' while a DMP is closer to 'anonymous, aggregate ad targeting.'
A CRM is for 'sales and relationship management'; a CDP unifies data from multiple channels to build 'marketing profiles.'