Customer Data Platform (CDP): definition, uses and business case
A Customer Data Platform is software that combines customer data from many systems into a single persistent profile available to other tools. The article covers the CDP Institute definition, the differences from CRM and a data warehouse, a worked business case for an online shop and typical implementation mistakes.
Amounts and values in the examples are illustrative and show how the calculation works. In projects, we use the client's data.
Dominik Dębski · Lemur Media · · reading time: 4 min
Reviewed by Dominik Dębski, 1 November 2026
What is a Customer Data Platform?
According to the CDP Institute, a Customer Data Platform (CDP) is software that creates and maintains a persistent, unified customer record available to other systems.The category name was introduced by David Raab in a blog post in April 2013. In 2016 he founded the CDP Institute. A CDP pulls data from many sources, merges records belonging to the same person (identity resolution), stores history and makes profiles and calculated attributes available to other tools via API, queries or exports. The CDP Institute also allows for an architecture in which CDP services run on data in the company's data warehouse.
How does a CDP differ from CRM and a data warehouse?
| System | Main use | Users | Data |
|---|---|---|---|
| CRM | Managing relationships and the sales process | Sales reps, customer service | Contacts, sales opportunities, activities entered manually or via integrations |
| Data warehouse | Analytics and reporting | Analysts, controlling | Data from many systems, modelled for analysis |
| CDP | Customer profile for activation in marketing and service channels | Marketing, e-commerce, analysts | Behavioural events, transactions, consents, calculated attributes |
A CDP pays off where the same customer data needs to reach many channels quickly. A company that communicates with customers mainly through sales reps more often needs a better CRM and data warehouse.
Which use cases justify a CDP?
A CDP is justified by use cases that require an up-to-date customer profile in several channels at once and deliver a measurable increase in margin. Typical applications differ depending on the business model.
| Model | Use case | Impact measure |
|---|---|---|
| E-commerce and D2C | Customer reactivation, post-purchase recommendations, excluding buyers from acquisition campaigns | Increase in contribution margin versus a control group, media cost |
| Media and publishers | Reader segments for subscriptions and advertising, content personalisation | Subscription conversion, advertising revenue per user |
| B2B SaaS | Product usage signals passed to sales and customer service | Revenue expansion, reduced churn |
| Marketplace | Communication with buyers and sellers by stage of activity | Share of active buyers and sellers |
When assessing use cases, the company also checks whether it has the consents and legal bases to combine data from different sources into a single profile.
How do you calculate the business case?
The business case is calculated for specific use cases, based on the increase in contribution margin, and compared with the annual cost of the platform. Example: an online shop has 40,000 customers who have not bought anything for 6 months. Segmentation in the CDP raises the conversion rate of the reactivation campaign from 2.0% to 2.6%.
| Item | Calculation | Result |
|---|---|---|
| Additional orders in the campaign | 40 000 × 0,6% | 240 |
| Additional revenue | 240 × PLN 220 | PLN 52,800 |
| Additional contribution margin (30%) | 52 800 × 0,30 | PLN 15,840 |
| Per year, with 12 campaigns to successive groups of customers inactive for 6 months (40,000 on average in each) | 15 840 × 12 | PLN 190,080 |
| Annual CDP cost (licence and maintenance) | assumption | PLN 250,000 |
| Result | 190 080 − 250 000 | −PLN 59,920 |
A single use case does not justify the investment. The business case needs two or three use cases, e.g. reactivation, post-purchase recommendations and excluding existing customers from acquisition campaigns, each with its own calculation and owner.
How do you prepare for implementation?
- A list of use cases with an estimated value and an owner on the business side.
- An inventory of data sources and customer identifiers: e-mail, customer number, device ID.
- Rules on consent and the legal bases for processing for each application.
- An architecture decision: a standalone platform or services running on data in the existing warehouse.
- A pilot of one use case with a control group that measures the actual uplift.
A pilot usually lasts one or two quarters and ends with a decision to extend or close it. The control group is a randomly selected share of customers who do not receive communication based on CDP data. The difference in contribution margin between the targeted group and the control group is the only reliable measure of impact. Without it, sales growth attributed to the platform may be due to seasonality or parallel campaigns.
What are the common mistakes?
- Buying a CDP before defining use cases and their value.
- Calculating the business case from revenue instead of contribution margin.
- No control group, so the conversion uplift is estimated rather than measured.
- Carrying data quality errors from the sources into the CDP without fixing them.