Preparing your data for the AI enhanced future

By Morgan Templar, CEO, First CDO Partners

For the first time, maybe ever, Data is on the tip of every CEO and Board Members’ tongue. “Data is our Number One Priority,” is ringing through the boardrooms of countless organizations. The sad truth is summarized in this quote, “Everyone is Ready for AI, Except the Data.”

The more recent  of my books was published in 2018. It introduced the GOVERN framework. Five years ago, many of the topics covered were only beginning to become mainstream. I wrote the book because I could see the technical components of good data governance were more solid than the people aspects. Of course, we all know that if people are not convinced of the value of a program or system, they won’t invest their time. At its heart, “A Culture of Governance” is a book about Data Culture and change management.

It’s time to modernize our thinking and to step forward to today’s concerns and opportunities.  This article will share how the Grow component of the GOVERN framework applies today.

The GOVERN framework has 6 general topics. As you can see on the Left column, each letter stands for an aspect of expanding governance. The topics in the Right column align to today’s concerns.

GROW Expand Data Governance to Unstructured Data
OPTIMIZE & OPERATIONALIZE Preparing for AI/ML & Harvesting Value
VALUE Talking with the CFO about Data
ENRICH Addressing AI Challenges
REVISIT CORE PRINCIPLES Have a strong Data Foundation
NEUTRALIZE NAYSAYERS & NEXT Managing Naysayers & Planning for what to do Next


Data is both an Asset and a Liability. On the one hand, we use data to run our businesses, understand performance, provide inputs to strategy decisions, plan and implement marketing campaigns, and many other uses. On the other hand, most of our data is unstructured and/or unknown. Growing your program to include more of the organization’s data helps to mitigate these risks before they become more than a potential liability.

The difference between Information Governance and eDiscovery is a Lawsuit.

When organizations are subject to legal action, it impacts what data can be archived or deleted, and puts a fence around data defining it as on “Legal Hold,” sometimes for years. When the legal team subpoenas your data, they use system crawlers to pull everything, from documents and transactions to recordings of online meetings, to phone logs, text messages, shared drives, individual drives, and whichever system data is relevant. You may feel confident in the information managed in your systems. But are you as confident about the data swamp that exists everywhere else?

Data in organizations fall into three categories: Known Data (Governed), Known “Unknown” Data, and Unknown Data (Dark Data). According to Gartner, something in the range of 80 – 90% of corporate data is unstructured, putting it into the Known “Unknown” category or the Dark Data category. Known “Unknown” Data includes information in paper records, data on SharePoint, Slack, business databases such as Access, SAS or SQL and other known sources. Dark Data on the other hand is the information on people’s hard drives, print outs, binders of historical performance and exported content, etc. This unstructured data is discoverable during eDiscovery. BUT it could also be a treasure trove of information available for fact-based decisions, forecasting, and other value drivers.

Unstructured data is, without doubt, the biggest liability in your organization. This liability falls into three categories:

  1. Legal Risk – Whether you know where the data is or not, everything is discoverable.
  2. Financial Risk – You have value in data and information that is wasted.
  3. Cultural Risk – Institutional knowledge in the heads of a few Subject Matter Experts belongs to the company but needs to be harvested to create value. It must be documented and reviewed to ensure it is not lost to retirement or perpetuating poor data hygiene.

Beyond the Basics of growing your governance program, you need to look at all your data to decide what needs to be governed formally versus managed in routine system lifecycles. New opportunities exist in the areas of Analytics Governance, ModelOps, Intellectual Property management, and custom configurations.

  • Analytics governance: Tracking the models, why they are needed and how they are to be used. Know where they are being used and how to interact with them.
  • ModelOps: Manage model deployment and model monitoring; manage model versions; mange security of the models; document model compliance.
  • Intellectual Property: Information regarding patents, trademarks, copyrights, and other intellectual property associated with the product.
  • Configuration Management: Metadata related to product configurations, variants, and options.

Growing your program to include more of the organization’s data provides the opportunity to mitigate these risks before they become more than a potential liability.

We all want the benefits of AI, but first we must lay the foundation of Data Governance.

Hot Topics

Related Articles