On April 27, at the 2023 Data Governance New Practice Summit, Mr. Chen Lijie, a partner of KPMG Consulting (China) Co., Ltd., shared a keynote speech on "Data Asset Management and Operation in the Digital Economy Era".
The following is the transcript of Mr. Chen Lijie's speech. For the convenience of reading, the editor has made some wording changes and text optimization. good morning guys! It is a great honor to be invited to participate in the 2023 Data Governance New Practice Summit. Today I would like to take this opportunity to share with you some of the latest industry trends we have seen in the process of digital economy exchanges at the national level and consulting projects in specific enterprises.
1. The trend of factorization of data assets in the era of digital economy
In recent years, the National Development and Reform Commission, the Ministry of Finance, the Ministry of Industry and Information Technology, and the Cyberspace Administration of China, including local governments, have introduced a series of management systems and measures. Previously, we communicated with the High-Tech Division and the Price Monitoring Division of the National Development and Reform Commission. They received very positive feedback from many industry colleagues after the release of the "Data 20 Articles" . , revenue distribution, security governance four pillars , then how to implement? Are there implementation rules.
In this regard, we propose that it is not yet ripe to formulate relevant implementation rules. It is recommended to conduct pilot work first, that is, to select one or two industries and scenarios to implement these contents first . In terms of industry, choose industries with relatively leading capabilities in data asset management and operation such as finance and communications; in terms of regions, choose regions where the digital economy itself is developing rapidly, such as Zhejiang and Jiangsu; in terms of scenarios, choose industries that are more closely integrated with business Scenarios; through pilot work to promote the development of follow-up work.
It has been more than three years since the Fourth Plenary Session of the 19th CPC Central Committee first used data as a factor of production, and the relevant supporting policy and institutional documents have matured. Now the pilot work is being done, and comprehensive promotion will be carried out immediately after the pilot. Therefore, the trend that private data is the core factor of production must be irreversible. **Traditional data governance is mostly driven by regulation, especially in the eight years of financial regulation, the China Banking and Insurance Regulatory Commission has issued relevant compliance requirements. It is understood that in 2021 and 2022, the People's Bank of China and the China Banking and Insurance Regulatory Commission will impose more than 1 billion fines on financial institution data, including just after the first quarter of this year, and the fines have exceeded 600 million. However, with the further implementation of data elements, future data governance will gradually transform into digital-driven and value-driven, which is related to the core competitiveness of enterprises and is an inevitable trend of future development.
The "Regulations for the Administration of Regulatory Statistics" issued by the China Banking and Insurance Regulatory Commission in December last year specifically mentioned data applications. It is recommended that financial institutions should fully consider data applications and data-enabled businesses in the process of submitting regulatory statistics. That is to say, do not simply process and report data statistics, but also consider the business application empowerment scenarios within financial institutions, and use some latest technologies such as big data and artificial intelligence to empower business when necessary, which also reflects The development trend of digitization, intelligence, capitalization and valueization has been realized.
2. Allocation of data property rights
As mentioned earlier, the "20 Articles of Data" proposed four pillars, the first of which is data property rights . Now the state has downplayed the ownership of data, and enterprises have the right to hold it. By processing the data they hold into data products, they have the right to operate and operate. After providing it to the outside world, users have the right to use it. In this process, data property rights must be considered, which is the biggest difference between current data asset management and traditional data governance 1.0. If it is said that doing data governance is more about acknowledgment of responsibility and punishment for quality problems, then the current data asset management is more about confirmation of rights, and the distribution of benefits after rights confirmation is a very critical point.
In the process of doing data asset management and operation projects with some large banks recently, we have tried to put the valuation and pricing of data assets into the bank's internal performance appraisal, that is to say, if there is a business department claiming these data in the future, the responsibilities and rights peers. Once the data department and the operation department make these data into corresponding data product services, whether it is the benefits generated by internal empowerment, there will be internal transfer pricing, or if external empowerment can have revenue distribution, the business department will think that data governance Or data asset management is not a burden, because performance appraisal is not a simple punishment for poor performance, but a good internal fund transfer pricing, external direct income distribution, so this plays a role in the promotion process Great effect.
The principle of income distribution in "Data 20" is that whoever invests will distribute it, and departments that have not invested have no say in income distribution. At present, the thinking of the business department is also gradually changing. We found that in the past, data governance was dominated by the data department and the technology department. The technology department and the business department pull together to carry out pilot work, and then it can be fully rolled out.
3. Financial data asset operation and value realization
Combined with the background of the country's entire digital economy, how should financial institutions carry out data asset operation management and value realization? First, let’s talk about the overall framework. This framework is also based on traditional data governance and our recent emerging data market management . , Governance, Integration, Agility, Guidance, Evaluation, Appraisal, and Comment” totaling 12 management key points.
"establish"
First of all, "build", the enterprise must have a data foundation, and for these data objects, it must first establish a basic foundation, "build" includes "collecting-communication-remittance", mining refers to collection, from the perspective of business, what data is collected , what is the range? What is the standard? What are the requirements? The second is "communication". After mining, it needs to be integrated and connected internally, between departments, between systems, and with business scenarios and business processes. It cannot be said that data, business and technology are separate layers. The third "collection" refers to the convergence. Many business departments mentioned whether the data of the business departments can be collected on a unified data platform for unified management.
"Tube"
After the "building" is completed, it is necessary to consider "management". "Guan" has derived three words, "pan-gui-zhi". This is more traditional data governance. We often mention that we need to do an inventory first, classify and classify the data to form a catalog during the inventory process, set standards and norms based on the data asset catalog and then go to governance: on the one hand, standards and specifications are out of standard, on the other hand, quality monitoring, rectification and improvement. Then we have to consider "use".
"use"
** "Yong" includes "rong-min-yin". **I have been mentioning the need for pilots above, but in fact it is to find pilots in terms of business applications. First, there must be integration, including the integration of business applications for data governance that the People’s Bank of China has always mentioned. How to integrate? It is necessary to find some specific business scenarios, such as customer marketing scenarios, what data will be involved in customer marketing scenarios? How is the data quality? What about security compliance requirements? Can it be used? Do you want to desensitize, what kind of marketing data product to create, and what value and utility can this product play? It is necessary to plan out the entire integrated business scenario, and then develop and implement it agilely after planning, and integrate the concept and requirements of data governance with development through Data OPS. "Introduction" refers to guiding the business department to use data. The business department is very clear about the business scenario empowerment of the data, but what kind of technical means, what model, what algorithm to use, and which external organizations to find Cooperation requires guidance.
"Comment"
**The lowest level is "evaluation", and we have launched "evaluation, evaluation, and comment". What is more evaluation? It is the valuation and pricing of data assets mentioned above, which is the foundation and a supporting mechanism for the future. At the same time, there must be comments and comments, or what we call evaluations. From the perspective of the management, we have to evaluate who is doing well and who is not doing well. From the perspective of third parties and consumers, there should be comments: Is the data asset management good, and is the customer experience high? If the data governance is perfect, but the business department or consumers or users suggest that the experience is poor, then there is a problem. Therefore, "assessment, evaluation, and comment" are based on three different perspectives, the perspectives of producers, managers, and consumers, which basically constitute a set of evaluation systems. The above are the 12 core principles of data asset management in our entire enterprise. management points.
4. The trinity operating system of "catalogue-account-product"
The starting point of data asset operation is to form a unified data asset catalog through inventory, and it is recommended not to divide this catalog by traditional IT technology themes, but to take what the business department can understand. The underlying modeling can be done through the traditional modeling methodology, but since the presentation of the business is said to be thousands of people, then the data asset label can also be thousands of people, we can have data labels, technology labels, At the same time, it is also possible to create a business label for data assets for the business department, which is convenient for the business department to subscribe, query, and maintain, but it can be mapped and matched with the underlying model. Then confirm the rights, classify and classify, and finally create data products.
** For data products, we recommend that enterprises consider creating a trinity operating system of "catalogue-account-product" in the future. **All data that we consider valuable can be placed in the asset catalog, which is the foundation. In this asset, we need to consider the concept of an asset account. What is more about this account? Two aspects, one aspect is bound to the ownership of the asset, who owns the asset, and establishes an asset account. The second is related to secure access. Which data can be used and which cannot? Through data asset accounts to clarify and based on the results of classification and classification, which accounts can be accessed, which accounts can not be accessed, which accounts can be used and which ones cannot be used? Therefore, the account achieves two purposes, one is ownership and the other is security compliance.
5. Creation of data products
Regarding data products, I want to emphasize that financial institutions generally create data products by the technology department and data department, and they are more based on developmental thinking, such as developing a few reports, developing a few indicators, developing a cockpit dashboard, Developing a set of models, etc., basically stays on the traditional development concept, but the creation of data products we propose here needs to be converted into a product concept, and data should be regarded as a financial product, from planning, design, exploration to development , Deployment, covering the entire life cycle process such as growth period, maturity period, recession period, delisting and outage. If you want to go in the direction of assets, you must be inseparable from product creation and product operation. At present, many financial institutions are relatively lacking in terms of methodology, personnel positions, and talent allocation. In the future, we must gradually strengthen them.
6. Compliant operation of data assets
Security compliance is the bottom line. In the entire financial industry, relevant laws and regulations, industry terms and requirements, including some standards and regulations, we simply sorted out almost 100 policy documents at that time, among which the "Data Security Law" can identify the specific requirements, and these The requirements can correspond to 12 industry standards, 10 local regulations, 17 national standards, 2 national laws, and 17 ministerial regulations. At the same time, we need to string all these contents together and return them to the specific scenarios of our data management and operation, and we need to do fine-grained compliance operations for each legal clause to be followed in each scenario.
We now do data security management more directly based on the results of classification and classification to make security policies, and implement them through system control. At present, financial institutions have gradually begun to do detailed, item-by-item, and compliant searches, and the items after the search fall into specific data application scenarios, and can perform risk assessment and impact analysis on compliance for this scenario As well as the series of solutions that follow, it is of course very difficult to achieve this level of granularity, which is also our future development goal.
7. Data asset value assessment
Many industry topics are now studying the valuation and pricing of data assets. Whether it is from the perspective of inherent value, input cost, or future replacement value, factor multiplication benefits, and market value, a comprehensive valuation model must be formed. Once the valuation of data assets is implemented, I believe that many problems will be solved.
During the entire process of valuation, we also put forward a point of view. At present, more static valuations are used, including the attempts made by many financial institutions are also static valuations. In the future, if we really want to put the valuation into the whole life cycle of data products, then it must be a dynamic valuation, from data collection, processing, storage, packaging, product creation, and finally to the realization of transaction circulation to generate value, During the entire life cycle, each node must be evaluated, and after the evaluation, there must be a value-added link , so that refined management can be achieved, and it is possible to say who collected the collection list and how much was generated during the collection process. Value, what value is generated in the process of processing? In the future, we will realize the profit sharing of business departments. Without this evaluation system, we will not be able to share profits.
Due to time constraints, I will share the above content with you today, thank you for listening!