This note is copied from "CDP Use Cases: What Users Want", the report is researched by CDP Institute
- Building segments is the most widely required capability, across 72% of use cases. It is an option across all tasks except predictive models.
- The two core data assembly capabilities, ingest data (96%) and create profiles (90%), are used by nearly all use cases involving those tasks. ID resolution (75%) and clean data (71%) are listed notably less often, perhaps because a significant fraction of users already handle those functions outside their CDP.
- Predictive models (85%), attribution models (79%) and recommendation models (79%) are the most common capabilities among analysis and predictive modeling tasks. Even tasks such as creating a data set (56%) and appending model results on a recurring basis (58%) are less common, again perhaps because users have other ways to achieve them. Calculating scores in real time is the least required capability (42%), suggesting that it applies to a set of specialized projects.
- Building content (74%) and creating multi-step process flows (72%) are shared by the most outbound campaign, realtime interaction, and cross-channel orchestration use cases. Defining channels to use (78%) is common but applies only to orchestration projects. Sending messages for external delivery (61%) is notably more common than delivering messages from within the CDP itself (50%).