Datazoom’s Custom Data Management feature lets admins standardize, organize, and operationalize custom events and custom metadata collected from SDK implementations and, for Javascript collectors, from remote instrumentation. The goal is to turn raw custom keys or query selector configurations as part of remote instrumentation into custom aliases: data points that can be individually selected in data pipes, filtered, enriched, transformed, and routed consistently across the platform.

Core concepts

Unmapped keys

An unmapped key is a raw custom data point that Datazoom has seen from a collector but that has not yet been associated with an alias. In the side panel, unmapped keys are shown in italic text.

These are usually the starting point for governance. For example, if one app sends bitrate, another sends BitRate, and another sends btrate, each may first appear as a separate unmapped key.

Aliases

An alias is a standardized, user-defined representation of a custom data point. It groups one or more source keys under a single managed definition. Aliases appear in normal font in the side panel.

Using aliases gives you:

• Consistent naming across collectors and implementations

• Cleaner downstream schemas

• Better compatibility with filtering, enrichment, and transformations

• A single place to define meaning and data type

Primary workflows

1. Turn an unmapped key into a new alias

1. Open Settings > Custom Data.

2. In the side panel, select an unmapped key.

3. Review its details in the main panel, including which collector configurations are sending it.

4. In the Add to Dictionary section, choose Create An Alias.

5. Create the alias. The selected unmapped key becomes the initial Master Key and the first row in the mapped keys list.

This is the right option when the raw key represents a new business concept that does not already exist in your custom data dictionary.

2. Map an unmapped key to an existing alias

1. Select the unmapped key in the side panel.

2. In Add to Dictionary, click Map to Alias.

3. In the overlay, search for the existing alias that represents the same concept.

4. Select the alias and confirm.

Mapping is type-restricted. A metadata key can only be mapped to a metadata alias, and an event key can only be mapped to an event alias. Equivalent source keys from observed from different collector configurations are managed jointly (i.e., mapping operations for one, are applied to each equivalent key).

Managing aliases

When you select an alias, the main panel displays the alias’s details. Admins can edit the alias.

Alias Definition section

This section defines the identity and meaning of the alias.

• Alias Name: The display name used throughout Datazoom.

• Master Key: The key sent to connectors by default. It is required and must be unique.

• Definition: A plain-language explanation of what the data point means.

• Data Type: Required for metadata aliases. Available options are Boolean, Number, String, and Array of strings.

Defining metadata types precisely

For metadata aliases, type definition is more than labeling. It improves validation and enables richer downstream behavior.

For Number, users can define details such as units like milliseconds, bytes, or pixels, and choose numeric representation such as Int, Float, or Double. Decimal types can also include precision and min/max constraints.

For String and Array of strings, users can assign formats. If Value List is chosen, a Manage link opens an overlay where allowed values can be maintained.

Source Collector Mapped Keys section

This section shows every source key currently mapped to the alias.

From here you can:

• Unmap keys: Select one or more mapped keys and click REMOVE. The removed keys return to the unmapped state.

• Map new observed keys: Click ADD NEW and select existing unmapped keys in the overlay.

• Pre-map keys not yet observed: Use Create a new key to define a key in advance.

This is useful when you already know that another app, platform, or team will emit a variation of the same business field later.

Viewing change history

Each alias has a change log. Click the history icon in the upper-right area of the alias panel to switch to the history view.

The log shows:

• What changed

• Who made the change

• When the change was made

This makes custom data governance auditable and easier to maintain across teams.

Creating on-demand custom keys for Javascript collectors

Datazoom also supports on-demand custom data, also called remote instrumentation, for Javascript applications. This lets users define new custom events or metadata without shipping new application code. Please see Remotely instrument custom data for instructions.

How custom data is used in Data Pipes

Once aliases are defined, they become selectable in Data Pipes just like standard Datazoom dictionary points.

1. Open the relevant Data Pipe.

2. Go to the Data Points tab.

3. Find the alias under Custom Events or Custom Metadata.

4. Select the checkbox to include it for collection and delivery.

By default, connectors receive the alias’s master key. If needed, users can change this behavior per alias in the Data Pipe by hovering over the alias, opening the three-dot menu, and choosing the option to send the original key name instead.

Aliasing standardizes how data is managed in the console, while Data Pipe settings control what is actually routed and under which output key.

Behavior that users should understand

• Observed first, governed second: SDK-sent custom keys appear first as unmapped keys, then get normalized through aliasing.

• Events and metadata are managed separately: Their lists, mapping behavior, and allowed destinations are distinct.

• Custom events are prefixed downstream: Datazoom automatically prepends custom_ to custom event names when collected or delivered. Custom metadata does not use that event-name prefix because it lives as metadata rather than as an event type.

Recommended operating model

• Use clear alias names that represent business meaning, not source-system quirks.

• Keep definitions plain and precise so anyone can understand the field.

• Assign data types for metadata aliases as early as possible.

• Map naming variants into one alias instead of letting near-duplicates accumulate.

• Use remote instrumentation for fast Javascript experimentation, but govern those keys the same way as SDK-created keys.

• Review change history when troubleshooting naming or schema differences.

• After aliasing, explicitly enable the needed aliases in each Data Pipe.