Quality Standards

The Core Attributes of a
Data Product.

In order to be classified as a 'Product,' a dataset must adhere to strict foundational criteria. These six characteristics guarantee that data is valued as a top-tier asset, producing seamless benefits.

Data Product Attributes and Principles

The DaaP Baseline

For a Data Product to be considered authentic, it must demonstrate these six fundamental attributes in order to reduce customer resistance and enhance credibility.

Discoverable

Users should not need to depend on tribal knowledge when accessing data. Data products should be easily searchable within a centralized or federated data catalog.

  • Registered in a catalog
  • Rich, business-friendly metadata
  • Clear ownership attribution

Accessible

Also referred to as Addressable, the data product must offer a distinct, programmatic address for secure access to the information once it is located.

  • Unique Global Identifier (URN/URI)
  • Standardized output ports (SQL/REST)
  • Machine-readable endpoints

Trustworthy

To rely on the data without conducting personal validation checks, consumers need assurance of SLAs and clear lineage.

  • Published SLAs (uptime, freshness)
  • Automated data quality testing
  • Transparent data lineage

Self-describing

A product should be intuitive enough that no engineer is needed to explain it; schemas, semantics, and syntax should be integrated seamlessly with the data.

  • Attached semantic definitions
  • Sample queries and documentation
  • Version-controlled schemas

Interoperable

Standardization across the enterprise enables seamless integration of data from different domains.

  • Standardized naming conventions
  • Shared master data references
  • Polyglot data formats (Parquet, JSON)

Secure

Security should not be an afterthought; access control, masking, and encryption must be managed locally and overseen globally.

  • Role-Based Access Control (RBAC)
  • PII/PHI data masking policies
  • Automated audit logging

Why These Attributes Matter

Lacking these essential qualities, a 'Data Product' is merely a table thrown into a data lake. The aim of these six pillars is to eradicate any obstacles for consumers.

A Data Scientist or Analyst can quickly generate insights in hours instead of weeks when they come across a dataset that is self-describing, guaranteed to be accurate through SLAs, and accessible via standard SQL.

Shifting the focus from data movement to data usability and value, Data as a Product ensures that data is valuable and usable upon arrival.

The Consumer Experience Check

1
I found it instantly. (Discoverable)
2
I know how to connect. (Accessible)
3
I understand the columns. (Self-Describing)
4
I can trust it's up to date. (Trustworthy)
5
I can join it with CRM data. (Interoperable)
6
My access was approved securely. (Secure)

Is Your Data a True Product?

Utilize our assessment tool to evaluate your current datasets based on the six fundamental DaaP attributes and pinpoint any deficiencies in your system design.

Back to Ecosystem