Skip to main content

The need for a taxonomy

How the taxonomy tool can increase data quality

Updated this week

Why do we need a defined Taxonomy?

Modern marketing data comes from many different platforms, each with its own naming styles, structures, and inconsistencies. Without a defined taxonomy, this quickly leads to:

  • Inconsistent naming (e.g. UK, United Kingdom, GB)

  • Duplicate or overlapping campaign structures

  • Manual cleanup before reporting

  • Confusing dashboards that are hard to trust

A taxonomy solves this by introducing a shared structure and common language across all your media data.

When a well-defined taxonomy is implemented across platforms, you gain:

  • Standardisation – Consistent naming rules across channels

  • Comparability – Easier to analyse performance between platforms

  • Clarity – Data is structured in a way that makes sense to everyone

  • Reliability – Less manual fixing means fewer reporting errors

In short, a taxonomy turns fragmented platform data into organised, analysis-ready information.

How the Taxonomy Tool Improves Data Quality

The Taxonomy Tool is designed to enforce structure at the point of data creation, rather than trying to fix issues later.

It improves data quality by enabling:

1. Structured Naming Conventions

Naming conventions define exactly how campaign or asset names are built, using defined levels such as Channel, Region, Campaign Type, or Objective. This removes guesswork and ensures consistency.

2. Controlled Values

Each level can have managed values, reducing free-text entry and preventing spelling variations, duplicates, or incorrect formats.

3. Standardised Data Across Platforms

Even though platforms differ, your taxonomy creates a unified master structure. This makes cross-channel reporting accurate and meaningful.

4. Built-In Governance

Permissions, fixed values, and predefined rules ensure teams follow the agreed structure without relying on manual oversight.

5. Reporting-Ready Dimensions

Because naming conventions are built from structured levels, those same levels can be extracted directly into reporting as clean, reliable dimensions.

Did this answer your question?