Data quality in Data Activation

With Data Activation, the success of your setup depends on the quality of your data and the possibilities to reliably identify users. For example:

• To connect profiles to events, or events to profiles, an identifier is needed. For fully anonymous users, targeting is more difficult.

• Inconsistent or stale profile data results in inaccurate audience membership.

• Each platform has specific policies, ID requirements, and sync schedules that dictate functionality, so it's important to understand your destination's data flow.

It's important to define a single source of truth for any attribute to maintain quality and consistency and get the outcomes you're after.

When choosing which sources to connect and which actions to track, focus on your use case and start small. There are plenty of opportunities to expand later once you've verified that your initial setup works as intended.

Data quality key concepts

Note that not all data is ready for activation. Data that works well for analytics may fail when used for targeting or triggering journeys.

Before building complex use cases, validate that your data meets these requirements for the fields you plan to use:

• Consistent identifiers: Without stable IDs, profiles can't be unified or matched for use in destinations.

• Timely updates: Stale data leads to wrong targeting (for example, suppressing customers who have already churned).

• Accurate values: Incorrect attributes cause audiences to include or exclude the wrong people.

• Sufficient coverage: If most profiles lack an attribute, rules based on it won't activate enough people.

Identifier requirements

Every destination has requirements for the identifiers it can accept.

Identity coverage affects which destinations you can use. If a profile lacks an identifier that the destination accepts, you can't activate the profile to that destination.

Eligible vs. matched vs. activated

These are different concepts that affect how you measure success.

A profile can be eligible but not matched if the destination can't recognize the identifier. You can match, but not activate a profile if campaign rules, frequency caps, or consent settings prevent action.

Why match rate differs from audience size

Match rate is typically lower than the audience size. Common reasons include:

• Identifier coverage: Not all profiles have the identifier the destination requires.

• Destination identity graph: The destination may not recognize all identifiers.

• Data formatting: Hashing or formatting differences can cause match failures.

• Consent constraints: Some profiles may be excluded based on consent status.

When planning data activation, expect the match rate to be a subset of the audience size. The gap depends on the destination and your identifier coverage.

Diagnosing "destination received fewer profiles than expected"

Best practices for getting started with Data Activation

Start with what you need

Identify the 2-3 use cases you want to activate first. Then work backwards to determine which events and attributes those use cases require.

Connecting many sources before you have clear use cases creates governance overhead without the activation benefit. You'll spend time on data quality issues rather than on activation.

Know your source of truth

For any attribute, 1 source should be authoritative. If customer status comes from both CRM and billing, decide which wins and don't ingest conflicting values.

If "active customer" means different things in your CRM or in your billing system, your audiences' membership will be unpredictable. Align the definitions across all relevant systems before ingesting data.

Ensure identifier consistency

Every source you pull data from should include identifiers that can be matched to a common customer record. If different sources use different ID schemes without a mapping strategy, you'll end up with duplicate profiles.

Each source, using its own identifier, without a mapping approach, leads to fragmented profiles. Establish your primary identifier and ensure all sources include it.

Prioritize freshness where it matters

Depending on your use case, you need to define how fresh the data should be for your data activation flow and how often the data should sync.

• Real-time sources (web tracking) are needed when you want to react to behavior quickly.

• Daily syncs from CRMs or warehouses are often sufficient for status and attribute data.

• Match your sync cadence to how often the data is likely to change and how quickly you need to react.

Ingesting data once and never updating it creates targeting based on outdated information. Make sure that all your data sources have a defined refresh cadence.

Keys to successful activation

A Data Activation destination connector translates orchestration decisions into destination updates. Typical behaviors include:

• Add: When a profile qualifies for an audience or journey step, the connector sends an add update.

• Remove: When a profile exits an audience or journey, the connector sends a remove update.

• Sync frequency: Depending on the destination, updates occur on a schedule or near real time.

The connectors control the data transfer and list membership. They don't control what the destination does with the data after it arrives.

To activate your data successfully:

1. Provide the right identifiers. Ensure profiles have identifiers that the destination accepts.

2. Set up the connector. Configure authentication and identifier mapping.

3. Connect audiences or journeys. Specify which audiences or journey steps send to this destination.

4. Monitor match rates. Compare audience size to destination reporting to understand the gap.

To understand different connectors for data activation, see our data activation destination setup guides.

Common expansion paths

Once you have set up data activation and can see that it's running smoothly, you can start thinking of expanding the scope. This means adding more data points, more data sources, and refining the goals.

Teams typically add sources in phases:

More resources

• How to measure goals and outcome signals