CRITICAL DATA: WHAT? WHY? HOW?

My experience with implementing the “critical” data concept spans four years. In 2019, I shared findings in the article “Scope your data management initiative by using CDEs.” At that time, this topic got a lot of attention. However, lately, talking to many data management (DM) practitioners, I discovered that they still experience an uncertainty about the practical application of this concept. This article is a short version of the presentation I will deliver during the Webinar scheduled for December 21, 2022, and organized by Trigyan : https://www.dhirubhai.net/events/webinar-implementingthe-critica7008557399829622784/

In this article, I am sharing with you the answers to the following questions:

• ?????WHAT is Critical Data?
• ?????WHY should a company use it?
• ?????HOW can a company implement it?

WHAT IS CRITICAL DATA?

HISTORY

I have thoroughly investigated the history of this concept. The concept of “critical data” has been on the agenda of data management professionals for 13 years now. ?I found the first reference to the concept of “critical data” in David Loshin’s book, published in 2009. Later, in 2010, DAMA-DMBOK first edition used this term in the context of data operations management and data quality. ?In subsequent years, David Loshin and Rajesh Jugulum applied this concept in their books devoted to data quality. In 2013, the Basel Committee on Banking Supervision came up with the “Principles for effective risk data aggregation and risk reporting” and added some new features to this concept.

CONTEXTS

The challenge is that these sources had quite different definitions of critical data or did not provide any definition at all. ?The reason for this was that they considered critical data in various contexts:

• ?????IT operations.

In this context, critical data helps organizations prioritize IT work.

• ?????Data protection and security

The key focus is on the protection of personal data. In this respect, all data that is recognized as personal data is also considered as being critical data.

• ?????Data management

Critical data elements become important when you deal with data quality, master data management, and data governance.

• ?????Risk management

In this context, data is critical for managing risks.

So, let me provide a summary of different definitions.

DEFINITION

After thorough investigation and analysis of different sources, I came up with the following definition:

CRITICAL DATA IS DATA, WHICH IS CRITICAL FOR MANAGING BUSINESS RISKS, MAKING BUSINESS DECISIONS, AND SUCCESSFULLY OPERATING A BUSINESS

What does it mean for data management professionals? You can reach three obvious conclusions:

1.????????This concept can be applied to all data types, including master, reference, and transactional data. You even can think about metadata as being critical.

2.??????You need to start defining critical data at the data user side of data chains.

3.???????To define critical sourcing data, you will need to go back along data chains.

WHY SHOULD A COMPANY USE IT?

In this article, I only touch upon the data management context. In this context, the are two key use cases for critical data:

• ?????Prioritizing data management initiatives

Very often, people argue: is it prioritization or limitation? Assume a company has 10 000 data elements, of which 1600 are critical. Will you deal only with these 1600 or prioritize them as the first set of elements? In practice, you usually speak about prioritization, not limitation. Applying this concept, you can correctly scope multiple data management initiatives: data quality, master data management, and data governance.

• ?????Building business cooperation and partnering with other business functions

The concept of CDEs also has something in common with the concept of “driver-based” planning used in financial planning and analysis. What we, data management professionals, call Critical data elements, financial people call “business drivers.” Driver-based modeling aims to build mathematical models between the key operational metrics and projected financial outcomes, i.e., revenue and costs.

After some investigations, I concluded that critical data elements in data management and business drivers and KPIs in finance and performance management are similar concepts.

HOW CAN A COMPANY IMPLEMENT THIS CONCEPT?

To bring the concept into practice, you should follow the next steps.

Step 1: Identify the context

You need to know the reason why you are going to define critical data elements. As discussed above, different contexts lead to different definitions. Assume you will apply this concept in the context of a data quality initiative. In this case, you need critical data elements to prioritize building data quality checks along data chains, preferably from data sources to the point of data usage.

Step 2: Identify critical reports

The most critical reports for any company are financial and risk ones. Your next step should be documenting all financial reports for external stakeholders, like supervisory bodies. You should not be surprised if you come up with dozens or hundreds of such reports.

Step 3: Identify ?key performance indicators

Usually, a company manages its business using a set of key performance indicators (KPIs). ?First, you need to agree upon the set of KPIs to be taken into consideration. Then, you should discover which reports include these KPIs. It will let you limit the number of reports to be in scope. You can consider these KPIs as ultimate critical data elements (CDEs). By now, you have a set of reports and the set of CDEs. What is next?

Step 4: Analyze the quality of the ultimate CDEs

I added this step if you use the CDE concept for improving data quality. You need to know which ultimate CDEs have the most significant issues with data quality.

Performing Steps 1 till Step 4 can take several months, depending on the scope.

Step 5: Trace data chains back to define sourceable critical data elements

This Step is the most challenging task. The level of complexity depends on the complexity of your data chains. For the data quality initiative, you need to know physical data lineage that describes data movements and transformations between source applications and data usage points. If data lineage is not available, then some other techniques can be applied.

I will discuss the solutions to multiple challenges associated with the implementation of the “critical data” concept during the above-mentioned webinar.

About the author

Dr. Irina Steenbeek is a well-known expert in implementing Data Management (DM) Frameworks and Data Lineage and assessing DM maturity. Her 12 years of data management experience have led her to develop the "Orange" Data Management Framework, which several large international companies successfully implemented.?Irina is a highly in-demand international speaker and author of several books, multiple white papers, and blogs. She has shared her approach and implementation experience by publishing?The "Orange" Data Management Framework,?The Data Management Toolkit,?The Data Management Cookbook, and Data Lineage from a Business Perspective.

Irina is also the founder of Data Crossroads, a training, coaching, and consulting services enterprise in data management.?

To inquire about Irina's training or coaching, participating in your company webinar or event, please, email to?[email protected]?or book a free 30-min session at https://datacrossroads.nl/free-strategy-session/