Metadata Is Dead, Long Live Metadata

This article was written by Informatica and originally appeared on the Informatica Blog here:


Why the Metadata Market Has Shifted to a Data Platform Approach

In 2021, Magic Quadrant™ for Metadata Management Solutions was retired by Gartner®, which the company has been publishing annually since 2016. As a result, the 11 November 2020 edition marked the final publication of this report.

This move wasn’t a surprise to Informatica as we think Gartner foreshadowed it when it declared, “The metadata management market will cease to be a stand-alone market as all data platforms embrace advanced metadata management” in its March 26, 2021, report, “The State of Metadata Management: Data Management Solutions Must Become Augmented Metadata Platforms.”

Why Metadata Management Is no Longer a Standalone Offering

We agree with Gartner’s observation.

While standalone metadata management tools have provided value in the past, we believe they are not adaptive enough to meet the demands of today’s dynamic business environments. These tools suffer from the following disadvantages:

  • They may be disconnected from data consumers’ normal activities, workflows, and tools.
  • They ensure policy documentation rather than actual policy compliance.
  • The metadata is passive, tending to be document-centric at the expense of providing actionable intelligence at the point of use. By contrast, data platform–based unified and connected intelligence can make metadata active, enabling the business to focus on what matters most.
  • The metadata is not real-time. Metadata from various applications and processes must be extracted and added to the standalone metadata management tool, and as a result is outdated and incomplete.
  • Their business value and ROI are questionable due to a lack of clear alignment with data strategies and business objectives
  • They don’t promote metadata sharing between various data management applications and ecosystems.

To overcome these drawbacks, in 2017 we pioneered unified metadata intelligence powered by our AI engine CLAIRE to deliver contextual and actionable insights and recommendations across data management applications. “We were the first in the industry to introduce AI-powered metadata intelligence. This active metadata delivers actionable, contextual insights, recommendations, and predictions to help automate thousands of data management tasks,” says Informatica Chief Product Officer Jitesh Ghai.

The Informatica Intelligent Data Management Cloud (IDMC) and Active Metadata

IDMC is the industry's most at-scale, secure, and trusted data management cloud for digital transformation. IDMC provides hundreds of connectors and more than 50K metadata-aware advanced scanners to scan and ingest enterprise metadata about your business functions and their interlinks. Whether it is technical metadata, business metadata, operational metadata, or behavioral metadata, IDMC enables a comprehensive, unified enterprise metadata foundation across all data management applications.

This breadth of metadata is organized in the form of an enterprise metadata knowledge graph, which shows how data is interconnected and makes it easier to perform machine learning for inference. The metadata knowledge graph delivers consistent and connected data intelligence across all data management applications.

IDMC enables a comprehensive, unified enterprise metadata foundation across all data management applications.IDMC enables a comprehensive, unified enterprise metadata foundation across all data management applications.


Activating Metadata for Automated Outcomes

IDMC delivers the most intelligent and efficient data management automation in the market by leveraging over 11 petabytes of customer-consented metadata from a broad range of organizations to train and power CLAIRE models. CLAIRE learns from this rich metadata and meticulously crafted algorithms to make intelligent recommendations.

These algorithms automate business rule translation into data quality rules, identify column similarity across thousands of tables, and perform mass data corrections to mass data quality assessments. They also infer data domains and recommend next best transformation.