The Era of Data Transformation

The Era of Data Transformation

Data, called the oil of the Internet era, is AI's priming water and the foundation of corporate digital transformation. In particular, in the digital era where everything goes up to the cloud, data is considered the last bastion of corporate competitiveness. How can we use data, which is so basic and at the core, to create a good outcome for a company's digital transformation?

Data transformation paradigm

The reason why Internet businesses were able to grow so much is that they have made good use of data. For example, what is the biggest difference between advertisements in newspapers or on TVs and those on the Internet? In addition, what is the biggest difference between using E-Mart and Coupang? And what is the difference between navigation and T-map? All of these differences lie in how data is used for services and businesses. Internet business evolves by measuring and collecting various data types and using them in businesses to plan and market products and improve business efficiency. It has a thorough data-driven decision-making and business systems.

Search ads enables marketing targeted only for those who are interested in a specific product and have a purchase intention. You can track who viewed the ads you released, how many people checked them, and what actions they took after viewing them. You can measure the effectiveness of the ads that way and use it to improve the next one. Coupang analyzes which products are sold most to whom in which regions and which products are purchased next by those who have purchased a specific product. It predicts which products will be sold to whom and where in the future, allowing you to purchase products in advance, store them in warehouses, and prepare for delivery. Since sophisticated consumption predictions are possible based on this data, fast delivery is possible, and cost efficiency is sought by not stacking inventory in the warehouse. T-map collects and analyzes the routes of drivers and vehicle speeds, informs them of the fastest real-time route, and predicts the time it will take to move to their destination by the day and time.

Fintech companies, such as Kakao Pay, Toss, Banksalad, and Lendit, collect and analyze financial data to recommend optimal financial products for users or provide financial services for investment techniques. The reason why data can be used for business innovation in various industries is that the two technologies are supported.

First, collecting data, which was previously difficult to measure, has become easier. With the development of smartphones, various Internet services, and sensors, sophisticated data measurement and the technology to collect and store such measured data in the cloud have been developed.

Second, big data analysis technology that can efficiently analyze accumulated data has been developed. In particular, the development of artificial intelligence technology is improving the quality of data analysis.

Autonomous cars could be developed to the level of commercialization because cars are connected to the Internet, vehicle driving information is accumulated in the cloud, and artificial intelligence that can operate vehicles automatically is evolving. When data is poured in continuously, artificial intelligence self-learns and evolves into better performance, so its performance gets better and better.

For this reason, data is the most important core and the first step in a company's digital transformation. Whether using AI, cloud, or any other technology, the core of a company's DT competitiveness is data. The core of DT depends on how much and where to collect data, how to analyze it, and how to use it for business. Therefore, the first step in the DT implementation process is tracking what data our company has been collecting. We need to diagnose what data we've accumulated in corporate management activities, where and how we’ve accumulated it, and how we’ve used it.

  • IoT - (Data -> AI : cloud)
  • Core tech tool
  • DT
  • Corporate management innovation

At that point, you can face up to your regrets and limitations, and check what needs to be improved in the future. To improve this, we need to organize what additional technologies are needed, what are the limits of the data we have now to apply such technology, what data we need to secure more of, where and how much we need to store the data in the future, who will analyze it, and how it will be analyzed. In fact, the data that a company has collected and the data that needs to be additionally collected in the future are a company's competitiveness that other companies cannot imitate. While other digital technologies can be copied equally by any company, data provides a differentiated competitiveness that is unique to the company. As such, before DT implementation, a company should be able to establish a strategy for data collection and analysis so that it can be used for DT advancement.

If you are a recording agency that produces music and provides it to fans, or a manufacturer that makes snacks and sells them to consumers through supermarkets or convenience stores, you need to know where, when, and how customers who consume music and snacks purchase and consume these products. In a word, you can sell more products only when you can collect and analyze customer data on who consumes what and how. Of course, this data also plays an important role in planning and marketing new products. If there is no such data or if it is not analyzed properly, it will be difficult for the company to survive in the future, let alone achieve sustainable growth.

The beginning of DT, DDDM

In the digital transformation process, technology plays an important role as a lever. The technology to be applied varies depending on the purpose, the area, and how the company innovates its business. In general, technologies, such as AI, blockchain, cloud, data and edge computing, 5G, IoT, metaverse, and NFTs, are technological tools used in digital transformation. Among them, cloud and data are the most universal and essential, and when the use of these tools becomes more mature, the introduction of AI will naturally be considered. If data analysts analyze the collected data one by one each time, it takes time and money, and subjective opinions rather than objective facts may intervene, but AI derives accurate analysis results at a low cost in that respect. Moreover, AI's analytical skills get better over time, which is a good thing.

The accumulation and analysis of data in the course of a company's management activities and its use in decision-making for important business is called DDDM, or Data-Driven Decision Management. However, as technology advances, such data is getting better and better in both quantity and quality. For this reason, some people refer to DT, which means digital transformation, as Data Transformation. As such, data is discussed as the most important method of digital transformation. Therefore, it is no exaggeration to say that DDDM is the most important and fundamental when companies innovate digital technology. DDDM refers to a company's operating system that systematically judges by analyzing data rather than intuition or personal judgment of leaders in the decision-making process. It is the most universally applicable and practical method to achieve results in a company's digital transformation process. However, in order to do so, it is necessary to establish a system that can systematically collect data within the company and have the competency to analyze the accumulated data professionally. In addition, based on the analyzed information, it is necessary to have the management leadership to use it to solve important business problems and actively introduce it into decision-making. Such an overall system is DDDM.

The biggest reason why DDDM has become important is that the spread of the web in the 2000s and mobile technology platforms in the 2010s made it easier to collect data for customers, and the accumulation of more sophisticated and diverse data in a company's management activities created an opportunity to make data-driven decisions more frequently and accurately. As more devices, such as smartwatches and cars are connected to the Internet following computers and smartphones, more data is being accumulated in the cloud. As more data accumulates in the cloud, it is possible to analyze it effectively with artificial intelligence and use it as a reference for better management activities. We will learn how to understand data and how to achieve effective business innovation through DDDM in the era of data.

Organizational operation for data-driven decision-making system

In this way, there are increasing cases of collecting and analyzing data that was difficult to measure before and using it for business. Beyond marketing, distribution, and finance, data-driven decision-making systems are recently being established in various fields, such as medical care, manufacturing, agriculture, and forestry.

However, in order to have such a data-oriented business system, the company's system and culture must be supported in addition to technology. Even old traditional companies are already accumulating a lot of data in companies. There is data in intranets, such as ERP, CRM, and SCM, and IT systems within various companies. The problem is that the data accumulated in this way is not systematically collected, and the reference information is inconsistent, making it difficult to analyze it. In addition, the analysis results are not used effectively for decision-making in actual business, so analysis and decision-making are often made separately. As a result, traditional companies recognize the importance of data but struggle with how to analyze and utilize this data.

Successful data-driven decision management (DDDM) is accompanied by changes in working culture, such as breaking away from stereotypes and innovating the decision-making processes in addition to technology investment
  • After clarifying the purpose of data utilization, define reference data > Discover small and diverse data analysis tasks → Trial and error > Invest in technologies, such as cloud, AI, which are IT infrastructure >
  • Establish a data governance and consultative body to bridge the gap between business sites ↔ Technology, top ↔ Bottom >
  • Apply it to efficiency by process and business function >
  • Expand areas by applying it to products, new businesses, and BM
  • Establish a working culture through company-wide training/campaigns
  • Data measurement > Collection > Accumulation > Analysis > Utilization

It is impossible to have a data-driven system only with technology. Technology is only a tool and, in the end, if a system for using this tool is not established, it may lack unity. Therefore, it is important to have an organizational system, process, and decision-making culture that can utilize technology in the process of implementing DDDM. In this respect, interest in agile organizational systems is growing.

Agile organization refers to an organization system that operates business promptly and is a method that was mainly applied to development work that involves programming. This organizational system has been mainly applied to start-ups that require fast work momentum and Internet companies where technological innovation is an important success factor for business. More recently, however, agile organizational systems have permeated smokestack industries, such as finance, manufacturing, and energy, as well as large companies. Why is interest in agile organizations soaring?

All business changes are made by people. In a company, people unite into organizations, and people's performance varies greatly depending on the organizational structure. An organizational system that can quickly catch market changes and competitors' movements and respond quickly allows people to gather their wisdom and execute it agilely in an era of rapidly changing technology. Field-oriented and customer-oriented thinking is required to foster an agile business system and decision. Data-driven decision-making supports this, and agile organizations are attracting attention as the organizational system that makes this possible.

An agile organization is a structure in which a single team that can handle tasks in a self-completed manner centered on tasks is gathered without dividing departments based on tasks. Even if there are several people in charge of planning, development, marketing, and operation, or even if developers with a single job are gathered together, it is possible to increase work concentration and work quickly with the same idea by allowing the team to perform a specific task on its own.

Importantly, a team composed like this needs decision-making standards to perform tasks quickly. The standards are based on the data collected from customers in the field. It is necessary to make hard-headed judgments based on data. Decisions should be made based on customer data, not on the personal taste or experience of team leaders or executives.

The existing business operating system is a method of establishing a strategy by conducting market research, obtaining approval from the supervisor for reports, and allocating budget and manpower to promote work. This decision-making structure not only takes a long time, but it is also highly probable that poor judgments can ruin the business. Agile organizations already have a complete understanding of the market and customers within the team, so there is no reason to conduct market research and to report because they can make their own decisions. The budget and manpower are already authorized to be handled within the relevant team, so the communication that occurs during the approval process is skipped. However, in the business process, decisions are frequently made based on market and customer data on the service strategy, product planning, and marketing as a whole.

This organizational structure is not easy to apply to existing companies composed of task-oriented hierarchies. You can try it for some projects, but it is difficult to apply it to enterprise organizations, and even if you do that, it is not easy to get results quickly and work well. Just because an organization is structured like this does not mean that the members belonging to it do their work autonomously. This is because people who have been immersed in the corporate culture of the past are unable to set responsibilities and authority on their initiative in the new organizational system, make customer-oriented decisions that meet the market's eye level, and perform business processes due to their past habits.

Therefore, the agile organizational system should not be suddenly applied to all work areas to achieve short-term results. It should be applied only to new innovative projects or tasks that can perform a single task based on a clear goal in the short term, and a new change management method must be found in the experience of overcoming problems. Thus, it should be small enough to tolerate failures in this process. If you take on the risk of failure by doing this on a project that is too big, you may not even want to try similar challenges in the future, so you should apply them to tasks within the scope of allowing failures and experience success and failure.

Furthermore, these organizations must operate flexibly so that they can be dismantled and merged freely at any time. A single team composed of agile organizations should be able to be easily dismantled and put into another task based on work performance or after the task is completed. In other words, it must be able to gather and disperse freely like an amoeba. With such an organizational structure, it becomes possible to focus on the work itself without falling into organizational egoism.

In addition, the strengths of the agile organizational system can be properly demonstrated only when the output of the project is checked frequently, verified from a customer- and market-oriented perspective, and improvement measures are derived. Instead of launching the final version completed after a long period of development to the market, the way to maximize an agile organization's performance is to divide the process into small tasks so that each product can be tested and verified, and to improve quickly by checking the response of these results. The quickness of being agile is not just about making it fast, but it also creates optimal efficiency in improvement by collecting and analyzing data on market responses in between.

IT companies and startups that promote business innovation surprisingly quickly form their organization in this way and conduct business centered on customers. Such representative companies include Toss, Baedal Minjok, and Kakao Bank in Korea, and Google, Spotify, Netflix, Alibaba, and Xiaomi overseas. In addition, among large corporations, this organizational system is being introduced, and companies using the system include Orange Life, Allianz Life Insurance, Hyundai Card, Chinese home appliance maker Haier, and Japanese electronic device maker Kyocera.

In order for smokestack companies and large corporations to avoid misuse of agile organization, they should not introduce such organization unconditionally into the system, but refer to the points discussed above and refine them into a system that suits their organization. In the process, it is necessary to analyze the data collected in the field and draw implications to use as the basis for decision-making.

Brightics analytics and data platform

So let's take a look at what data is being collected and where, and who analyzes it and how this data is used in decision-making. In general, data is not collected in one place but is gathered in a free-for-all, and analysis is only aggregated and analyzed by gathering data at the level of financial statements, such as sales and profits, and it is not being used for important decision-making. In addition, there is no data analysis team or data scientist who specializes in data only.

If we recognize the importance of data, we need to establish a strategy for how to collect and analyze it for our own DT to solve our business problems in the future. That is the beginning of the DT strategy. That is the data strategy. In fact, the most ideal thing is to analyze the data through individual dashboards in all business departments based on the data being archived in a single system, find the implications needed for the business, and use it for decision-making.

Moreover, a separate data analysis team is set up to enable neutral and objective data analysis reports on decision-making matters required for important business activities. However, having such a system and organizational structure requires considerable investment and capabilities unless it is a professional IT company. Therefore, these days, if you operate the company's overall system based on cloud, data is automatically accumulated to help you easily perform various analyses. During the process, AI automates and streamlines data analysis. In this way, cloud and AI play a big role and help in data analysis.

As such a solution, Brightics Analytics Services, an analysis platform optimized for businesses, and Brightics Data Preparation, which enables intuitive collection and management of data collected during business activities in one place, makes it quick and easy to introduce AI-based analysis services for business innovation.

Architecture - An AI platform optimized for enterprises, Brightics AI
  • Click each Brightics AI module to view detailed information
  • Data collection and processing / AI-based data analysis platform / Professional analysis service
  • data preparation > / machine learning > / deep learning > / analytics service >
  • data hub >

In fact, a company's DDDM process requires a system that can store, integrate, and manage data based on AI. As such a system, Brightics solutions allow enterprises to choose and introduce only what they need. Above all, these systems have two advantages: One is that there is no burden of excessive initial investment because it can be used and paid for only as much as necessary on a subscription basis, and the other is that high-quality results can be seen with an easy and convenient visual tool thanks to excellent AI algorithms.

Samsung SDS Brightics Analytics Enterprise AI analysis service that innovates business

Although there is a need for DDDM, it is easy to hesitate due to complex technologies and burdensome system introduction costs when trying to apply it to a business site, but with this solution, it is economical because it can be selected and used as much as necessary.

▶   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.

Jihyeon Kim
Jihyeon Kim

He’s interested in and studying how technologies make changes in our daily lives and society, and how they can be used for BM innovations in companies.