A data management platform (DMP) is software for building and managing audience profiles and segments. To understand a proposal, follow the data entering the platform, the rule defining the group and the system expected to receive its output.
Start with the data entering the platform
Adobe Audience Manager’s data guide lists first-party inputs from a company’s own customer interactions, second-party inputs from strategic partners, and third-party inputs from data providers or exchanges. Collected data is mapped to traits: defined attributes or conditions derived from the incoming data. Those traits give the audience rule something specific to test.
A segment expresses a qualification rule
In Adobe’s Segment Builder, traits can be grouped and combined using conditions such as AND, OR and AND NOT. A segment therefore has criteria you can inspect. Read those criteria alongside the source information; a convenient audience name is not a substitute for knowing what qualifies someone for the group.
Check the outbound connection
Adobe defines a destination as a receiving system, such as a DSP, ad server or ad network. Its documentation says delivery depends on what the partner can receive; technical constraints can rule out a transfer method. Segment Builder treats destination mapping as a separate, optional step. Building a segment and connecting it to a receiver are different pieces of work.
A fictional audience proposal
Imagine a fictional cycle-accessories retailer proposing “lighting-category visitors without a recorded light purchase.” Category-page events are available, but the purchase export has not been checked. The team can inspect the visit rule while leaving the purchase exclusion unresolved. Calling the group “nonbuyers” would claim more than these inputs support. Before sending the audience to an ad platform, the team would identify an existing destination. If that mapping uses a segment value, the team would check which value it expects. This is a proposed setup, not observed software output or a campaign result.
DMP, CDP and DSP describe different jobs
The CDP Institute’s definition emphasizes a persistent, unified customer record accessible to other systems. It also allows external sources and anonymous individuals. Combined with Adobe’s first-party input support, that makes “DMP equals anonymous, CDP equals identified” an incomplete distinction. Ask whether the task is to maintain customer context or define an audience, then examine the proposed system’s actual responsibilities.
A DSP serves the buying task: Amazon’s product documentation defines it through programmatic purchases of ad placements. A segment sent to a buying platform is an input, not itself a purchase of advertising. Keep the audience definition and the media commitment separate when reading the proposal.
Settle the intended use before export
Name the person who owns the input definitions, the audience rule and the receiving connection. Keep data-use approval separate from technical configuration; ask the responsible owner whether the proposed data and purpose have been approved. A working connection does not answer that approval question. If a source or mapping is unresolved, identify that specific gap before presenting the audience as ready to use.