Does Your Company Have a Data Culture?

Does Your Company Have a Data Culture?

The data culture is a comprehensive concept emerging everywhere the importance of data is discussed, such as data organizations and data-driven decision processes, and is the most important topic for all data leaders responsible for managing data.

The data culture values the use of data to improve decision-making and supports the actions and beliefs of members who believe that this decision-making process is important. As a result of these actions and beliefs, data naturally permeates organizational operations and mindset. Data culture aims to solve complex business challenges by providing all members of companies and organizations with the insights they need for a true data foundation.

According to a Gartner survey1, data culture appears to be a top priority for chief data officers (CDOs). A report2 from McKinsey, a management consulting firm, states that "data culture is decision-making culture.” In addition, the Data Culture Report3 from Alation (, a product vendor specializing in data catalogs, shows that most companies are investing heavily in prioritizing data within their organizations. More than 86% of companies have a chief data (analytics) officer, and 78% have a company-wide initiative for data foundation. Companies focus on fostering data culture using increased collaboration between business and data and analytics teams (44%) and data governance management at the point of use (41%) as the top two initiatives. In the case of top data culture companies surveyed, the figures are increasing to 59% and 56%, respectively, indicating that they are putting more emphasis on eliminating data silos.

percent who have a c-level data officer(n=300)
  • any : 85% / none : 14%
  • 44% have a CDAO, 23% have a CAO, 19% have a CDO
percent whose company has an initiative to become a more data-driven organization(n=300)
  • yes : 78% / no(net) : 22%
[Figure 1] Chief Data Officers and Company-wide Initiatives within Organizations (Source:

Since data culture has become a hot topic in all companies, I would like to accurately define data culture, examine its benefits, and think about how to build data culture.

What is data culture?

In short, the data culture is a corporate culture that makes decisions based on data. Companies try to build a data culture because they want to make better decisions. Before explaining the organizational culture called data culture, let's look at the 'consensus' and 'class' cultures that typical corporate organizations have for decision-making.

The ultimate goal of consensus culture is to “achieve consensus.” While a consensual culture can be felt as a good culture on a daily basis because everyone has a voice and no one's comments are ignored, it has the disadvantage of leading to slow decision-making and passivity in innovation. Due to the great pressure that everyone should come to an agreement, large organizations are less familiar with innovative ideas that make changes in existing directions.

It is class culture that stands at the opposite point of this. It values position or seniority above all else, and everyone follows the opinion of the person at the top. Since the ability to generate and select ideas is limited to a select few, the morale of other members who are not given the experience of challenging new markets is degraded. Such a culture is likely to ultimately eliminate creativity and miss opportunities for new changes. In addition, the decision of these “upper-class” people is likely not the correct answer in a situation where the market changes quickly and new markets are emerging.

In the data culture, the evidence of data is of paramount importance and has the highest value. It doesn't matter who speaks and what position they hold. What matters is whether the decision is based on data. Consensus and class cultures are quick and easy to use. You can always ask your colleagues for their opinions in the conference room or contact your supervisor with just one phone call. However, to build a data culture, that is, to be able to find answers from data, it is necessary to have a number of complex processes and systems that make data instantly available, reliable, and interpretable.

Organizations with a data culture are more likely to make the right business decisions than slow-moving consensus cultures and unconditional class cultures. According to Forrester Research4, organizations with an insight-driven data culture are about three times more likely to achieve double-digit growth than organizations that do not. Organizations with a data culture can make better decisions faster. According to the 2020 MIT CDO and Information Quality Symposium, companies with a data-driven culture achieve “increased revenue, improved customer service, best-in-class operational efficiency, and enhanced profitability.”5 Data culture organizations also have the additional effect of helping recruit and retain the best employees. More competent employees like working in an environment where logical data makes decisions rather than in a culture where internal politics dominate or implicit agreements are pursued. Finally, a data-driven culture increases the loyalty needed to manage an organization. In an authority-driven culture where the direction of work is executed according to the instructions of the upper person, the attempt fails if the consent and support of the members who back up the various new challenges are not obtained. By sharing data from a decision-making background and analysis results, data culture organizations can dedicate themselves to executing corporate plans and increase the chances of business success.
To build a data culture with these advantages, organizations must pursue three capabilities:

The first is data search and discovery. This means that organizational members must be able to find relevant data in a timely manner when they need to make a decision.

The second is data literacy. Members must be able to correctly interpret and analyze data to draw logical conclusions.

The third is data governance. Business organizations must establish and manage guidelines and standards to ensure that data is properly managed and that employees can use it correctly.

Let's take a closer look at each of these.

Data search and discovery

According to IDC, more than 64 zettabytes of data were created, captured, and consumed worldwide in 2020.6 Not only the amount of raw data but also the types of database systems for storing all forms of data, from data lakes to cloud data warehouses, are increasing simultaneously. In the past, companies did not have enough data to make the right business decisions, but organizations today have overflowing data. In other words, if the old question was "Is there data?”, now the question for data culture is "How do I find the right data?” Moreover, considering that the data you want can exist in many different systems and in many different formats, the right question might be, "How do I find the best data available?”

Data search and discovery is the feature that enables users to find, understand, and trust the information they need to make data-driven decisions. Data search and discovery is the basis of data culture. If people can't find the data they need, it doesn't matter if data governance is established in the company or if there are employees with excellent data utilization abilities. Data search and discovery needs to be broadly defined. It should enable users to find, understand, and trust a wide range of information assets, including raw data as well as unstructured information data, local text documents, and undocumented processes. Understanding data means an effort to know more about data assets. For example, what is the meaning of the data you use, where does it come from, who manages it, who else uses this data and what data do you combine it with for added value?

Trusting data means knowing if and how a given data asset can be used. Reliability is achieved only when you know if this data is accurate and most up-to-date, who can access this data, and what data policies are needed.

There are three common types of users for data search and discovery:

Data scientist: Modeling, advanced analysis, or data reprocessing
Business analyst: Finding added value through prepared data sets or BI outputs
Business user: Creating BI outputs to answer everyday business questions
In addition, this form of using data provides a greater advantage due to the virtuous cycle effect of mutually complementing and rising.
• The more people use it, the better the system becomes. It provides knowledge and leaves experience in relevance of usage patterns.
• The better the system becomes, the more people use it. This is because new users automatically flow in when the quality of content and search results improve.

With this virtuous cycle momentum, the power for data culture becomes better.

Data search and discovery system is also utilized for data governance to ensure proper management of data and compliance with policies.

For example, a data privacy manager can find all use cases of phone number data and label them as personally identifiable information. Systems built specifically for data governance tend not to satisfy analytical end users, but typical systems built for information seekers make data governance much more efficient. This is why data search and discovery systems are the basis of any data culture.

Data literacy

Data literacy is the ability to read, work with, analyze and argue with data. A step-by-step approach is needed to understand this.

3 steps of data literacy

The first step is “data analysis.” Data literacy begins with the ability to analyze data using multiple technologies and tools. Even this step requires many skills and competencies, but they are only the basis for overall data literacy. In other words, it means that these analytical skills and data access competencies are essential to starting a data culture, rather than thinking lightly and simply about the necessary tasks to build capacity at this step. If this step is not followed faithfully, visualization using data sets can lead to incorrect insights.

The second step is “drawing conclusions.” The next step in data literacy is drawing accurate conclusions from data. You should be able to draw the right conclusions to your business questions by choosing the right solution for your situation among various data sources and analytic technologies. This requires determining which technologies are best for a given type of question, understanding the limitations of data and analysis, and recognizing and alerting to numerous forms of cognitive bias that can interfere with rational reasoning.

Next is the “ability to persuade stakeholders at the top” of the data literacy hierarchy. Even if you have excellent analytical skills and cannot persuade stakeholders, the insights created by that data will lose their value. Persuasion using data starts with the step of properly visualizing data, makes a data-driven design, and becomes effective when telling stories using data is completed. To persuade others, you need to know the information clearly and provide reliable and convincing explanations. In organizations where methods of persuading stakeholders with data are established, factors that hinder organizational productivity, such as skepticism based on personal experience or groundless distrust of data, can be minimized.

Meaning of data literacy and how to acquire it

Data literacy is about enabling and empowering people not only to argue about data but also to feel free to argue against false claims made using data. In a data culture, data literacy is also important in the sense of resolving the information asymmetry between members who have access to data within the organization and those who do not.

The real goal of data literacy within the enterprise is to make data accessible to everyone so that all departments can use data to make better decisions for organizational success. Rather than simply reading numbers, graphs, and charts, it makes it easier to consume data by inferring useful meaning from various visual representations. If your goal is to help everyone in your organization become more familiar with data, let's look at how to help them and what the key elements are.

1. A data leader is necessary

Find people in your organization who will be strong data leaders and a group of members who can help create data literacy programs. This group is a team that can advocate the benefits of data literacy to those who are unsure of the value of data. It plays a role in identifying groups within the organization that are missing opportunities due to a lack of effective use of data and prioritizing data literacy programs.

2. Make a change

A transition to make all cultures of the organization a data-driven requirement should begin. The culture changes when company leaders regard data-driven decision-making as a model and require team members to do the same. It is essential to actively share and promote data success stories within the organization to understand the power of data, and it should be easy to see how data affects actual business outcomes.

3. Decide on systems and tools to access a wide range of data

Without a system established to effectively share data across the organization, data scientists, and analysts will face a bottleneck in increasing work efficiency. Introduce methods and tools for all employees to access data. Regardless of the tools, you need to provide a system that allows non-technicians to manipulate data and extract the most important information for themselves.

4. Invest in employee training

Giving all employees access to data is definitely an important step in data literacy, but it does not help much if employees do not use the data. Employees should understand how important data literacy is to the company's success. People who understand how important data is to achieve business goals become active in further training on how to interpret the data. Training in critical thinking and data technology is an essential factor in maturing an organization's data culture. All employees should be able to ask the following easily and naturally:

  • Data collection method
  • Value to learn from data
  • Reliability of information

Data always represents facts, but there are many cases where bias can be introduced in its interpretation and analysis, the data is misrepresented, the sample size of the data is too small to make strategic decisions, or the data becomes unreliable. People trained to think analytically about data know how to ask these questions. In addition, regular training on the safe handling of data and its ethical guidelines is required to prevent data leakage.

5. Start small and make continuous improvements through feedback

Organizations that have successfully improved their data literacy recognize that this is the result of iterative processes. Prioritize which groups need training, and apply the experience of the first group to the next group. Get feedback from the group on what works and what doesn't, and then refine the process and training to make it more effective.

Data governance

Data governance defines how data is collected and used within an organization and addresses the following key questions:

  • How does the organization's business define data?
  • Where does the data exist?
  • How accurate should the available data be?
  • Who can use it?
  • How can we use it?

These questions provide information about governance rules and principles for maintaining the highest quality of data throughout its lifecycle. Data governance creates the foundation for improving data quality and accuracy. Gartner describes information governance as "the decision-making authority and final limit of responsibility to ensure the actions in the process of evaluating informative values, generating, storing, using, archiving, and deleting information.” This includes processes, roles and policies, standards, and metrics that ensure the effective and efficient use of information to help organizations achieve their goals”7.

In short, data governance prescribes human behavior that complies with regulations on data. Creating a framework for data-driven decision-making and setting clear expectations on how to work with data improve the quality of data-driven processes and build trust.

Benefits of data governance

The benefits of data governance include:

1. Better analysis: Since data governance improves data quality and discoverability, analytics professionals can find, understand, and analyze data more quickly by following regulations.
2. Consistent compliance: Data governance helps data users comply with regulations, reducing the risk of financial and reputational damage.
3. Improved data management: Governance increases operational efficiency by minimizing redundant tasks.
4. Standardized systems and data policies: By standardizing systems and policies throughout the organization, it creates an ethical awareness regarding data use among users.
5. Enhanced data quality: Compliance with data governance enhances the quality of data generated, which creates more accurate processes and higher reliability.

Data governance requires users to follow the same guidelines to achieve higher quality and accuracy and provides a flexible framework to respond to new laws and regulations regarding data use.

In addition, it sets similar data usage processes throughout the organization. This setting is important because different processes can result in different data quality and accuracy. When this process is properly implemented, a data governance framework not only unifies the behavior of members in an organization regarding data but also helps members understand the rules themselves. In other words, a transparent governance framework clarifies the reasons behind data processes. This transparency builds trust and allows people to understand the system, creating a unified consensus, which is a data culture, on data use within the enterprise.

Forms of data governance

Successful implementation of data governance is influenced by organizational culture. The three approaches to consider are as follows:

1. Command and control data governance (top-down)

  • Assign members to a data steward (a role that ensures data is used effectively)
  • Fast and compulsory governance
  • Preemptive purchase of governance tools

2. Traditional data governance

  • Users must use data only as planned and defined
  • Passive governance
  • Uniform approach
  • Innovation through governance is not a priority

3. Active data governance

  • Based on voluntary motivation
  • Inclusion of governance in users' daily activities
  • Distribution of decision-making rights
  • The belief that people can be more efficient and effective if they are recognized and rewarded
  • Establishment of a common trust model
  • Achievement of business performance while managing risks

One of the most effective best practices for governance is an active approach that engages data managers and demonstrates business value.

Data analysis platform

The level of data culture within enterprises is not mature enough worldwide. The main reason lies in the lack of data literacy mentioned above. Literacy shows its influence in actual business by passing through the collection and search stages and drawing valuable conclusions through data analysis by all members. Data literacy is not just a problem with the software, but appropriate software can certainly help support and increase data literacy.

In this sense, Brightics AI8, an AI and analytics solution provided by Samsung SDS, is a product with global competitiveness in this field, and I recommend many companies actively utilize it. Brightics AI is a comprehensive AI platform. It collects a massive amount of information, processes the information, quickly analyzes it with AI, and visualizes the results in an easily comprehensible way. In short, it helps collect, configure, access and enhance various forms of data assets scattered across the enterprise, and helps create added value by understanding data through analysis and visualization tools using AI and deep learning technologies. In other words, members of the company can collaborate based on data, enhance understanding, and use artificial intelligence technology to materialize data into a strategic asset with infinite value.

Composite map of Samsung SDS Brightics AI modules
  • data collection and processing, AI-based data analysis platform, professional analysis service
  • data preparation > / machine learning > , deep learning > / analytics service >
  • data hub >
[Figure 2] Composite map of Samsung SDS Brightics AI modules (Source:

The added value created in this way is fed back through collaboration with other members, resulting in more advanced data literacy, and serves as a platform that completes the entire data culture through data search and discovery and data governance.


It is clear that data and analysis play a pivotal role in digital innovation. However, the data itself is meaningless. It should be provided in the right format, at the right time, with the context needed to drive business results. Building a sustainable data culture requires data-driven behavior and reliability around the use of data, the ability to use it, and clear governance, rather than the choice of data or necessary tools.

Samsung SDS Brightics Analytics -
Enterprise AI analysis service that innovates business with pre-built analysis models verified in the field

For companies to truly innovate and utilize data, they must focus on their workforce and prioritize data in their decision-making. Since the data culture consists of actions and beliefs about data that permeate the way organizations operate, members of the organization can make better decisions that lead to valuable business outcomes when they are empowered by data.

1. Gartner, “Survey Analysis: Fifth Annual CDO Survey”, 26 Mar 2020
2. McKinsey, “Why data culture matters”, Sep 2018
3. Alation, “The Alation State of Data Culture Report”, Sep 2020
4. Forrester, “Insight-Driven Business Set the Pace for Global Growth”, Oct 2018
5. MIT, “How to build a data-driven company”, 24 Sep 2020
6., “IDC: Data Creation Hit 64ZB in 2020”, 25, Mar 2021
7. Gartner, “Information Governance”
8. Samsung SDS. “인공지능/애널리틱스, Brightics AI”

▶   The content is protected by the copyright law and the copyright belongs to the author.
▶   The content is prohibited to copy or quote without the author's permission.

Share Open