Podcast generated by Notebook LM
Phil Ren, Founder of Nocoly.com Limited
In recent years, many enterprises have been digitizing their business processes, with some excelling and others struggling. We attribute this partially to digital literacy, which allows enterprises to navigate the complexities of digitization with ease.
Recording
The first step in digitization is recording, which involves converting people, activities, and physical objects into digital records. For example, in sales activities, enterprises need to record customer information, the sales conversion process, quotations and contracts. In procurement activities, they need to record supplier information, the inquiry process, and procurement orders.
Ideally, all objects involved in an enterprise’s operational activities should be recorded. In fact, all stages of an enterprise’s operations consume and also produce data. So, how should we identify the objects that need to be recorded?
In the theory of Enterprise Architecture (EA), the value chain theory is widely used, It is an approach to enumerate all management activities within an enterprise. However, the value creation processes of different industries and enterprises vary, so we do not need to follow the framework shown in the following diagram strictly. For example, in the software industry, there are no stages involving materials and logistics.
Once you have identified your own value chain, you can use it to check whether data from each stage has been recorded. Please note that the data refers to unified records across the enterprise, rather than employees’ individual records.
The other goal of the value chain is to achieve higher efficiency, which means using lower costs and generating greater output per unit cost. Recording data around the value creation process is to enable the measurement and enhancement of efficiency. For instance, when communicating with customers, the objects we need to record include the customers themselves, the content of the communication, the frequency of communications, and the feedback received. These data can be used to analyze the effectiveness of the communication.
The example above is just a part of the business operations, which is not complex. However, in this example, some information might be overlooked. For instance, the marketing department may have failed to record complete batches of marketing activities or gather customer feedback. The primary reason for these omissions is the focus solely on completing operational tasks without considering the PDCA Cycle (Plan-Do-Check-Act). Therefore, in this stage, the ongoing analysis and improvement must also be taken into account.
Specifically, when recording, one can list all the people, objects, and activities involved in the business process. Taking procurement as an example, “people” refer to procurement personnel, suppliers, and their contacts; “items” refer to the materials to be procured; and “activities” refer to inquiries and sourcing (presented in the form of quotation and purchase orders). More requirements will be put forward on this in the subsequent section on “Structuring”.
I have proposed a simplified enterprise architect framework called RPIC, which is based on roles and traverses processes and information, enabling the recording of all contents.
Ideally, all the data should be managed in a database, which users could interact with through applications. However, the reality is that enterprises will not achieve full scale digitization at once. Nevertheless, even without an information system, recording itself holds value. For example, lots of managers use Excel to create various spreadsheets, which they utilize for decision-making, operational improvements, and data queries.
In addition to Excel, alternatives such as Google Sheets exist, and these online documents can be considered as a prelude to a more sophisticated digital system.
Due to the lack of constraints imposed by information systems, it is important to pay attention to the following issues when using these tools:
- One file or spreadsheet should manage one object. For example, order information and customer information should be recorded separately.
- One file or spreadsheet should record objects of the same type, rather than creating separate ones based on dimensions such as months or regions. Even if separate files are created, they should use the same structure to facilitate consolidating.
- Record details rather than summaries. For example, for inventory, we should record each stock out and stock in, rather than just the net stock.
If a business is able to neatly record all its operational data in spreadsheets, it has basically completed the first step towards digitization. Next, it will be less difficult to improve efficiency using digital systems.
Classification
The classification of enterprise data refers to setting dimensions to various types of data.
When setting the dimensions for enterprise operation matters, some are classified based on size, some are categorized based on the number of employees, and some are divided based on sales volume. Typically, once a dimension is established, it is not changed.
The design of dimensions takes into account both operational efficiency (matching different sales talents and processes for customers of different sizes) and business analysis (analyzing sales volume and purchase quantities of different products among enterprises of different sizes).
In summary, classifying data serves the needs of operations and analysis. Therefore, the design of these dimensions requires strategies from management and should not be overlooked or simply copied from templates.
When classifying, it means adding column(s) in an Excel file. These columns differ from regular text and numeric columns in that they typically consist of a predefined list of options. Only in this way can dimensions achieve Normalization. Additionally, using relational objects can serve the same purpose (this will be mentioned in Structuring later). However, if the dimension itself is based on time intervals or numeric ranges, there is no need to manually establish these dimensions. Because data analysis tools could provide Clustering for numeric values and time.
When using Excel to record, different spreadsheets are generally not linked. Therefore, to normalize the dimensions, it is necessary to predetermine input rules, or the dropdown list can be used to ensure the value uniqueness. In fact, integrity in dimensions is crucial for data quality.
However, we cannot fully rely on Excel to maintain data. To establish systematic records, we need to establish relationships between data. For example, the order details of a sales order should be linked to product lists. Next step, let us talk about structuring.
Structuring
In the field of enterprise data, data structuring specifically refers to the Relational Data Model.
The example mentioned in the previous section involves a sales order being associated with sales order details. Additionally, the sales order details form is also linked to the product lists, as a sales detail may involve multiple products. The data structure, which reflects data relationships and contains no redundant information, is known as a relational database.
We generally use an ER diagram (as shown below) to describe the relational database. Each rectangular represents a data object, where the rows represent attributes, including the dimensions we mentioned before. The connecting arrows between two rectangles represent relationships.
Some may challenge that spreadsheets can also establish such relationships using a cell reference. While this is true, spreadsheets cannot enforce standards. What if users do not follow the standards? Then the data quality cannot be reinforced. Furthermore, for complex businesses, spreadsheets are obviously not enough.
However, spreadsheets with clear structure and classification can serve as the foundation for digitization. This digital literacy can accelerate an enterprise’s digitalization and save costs. Because when an enterprise introduces a system, it can simply import these spreadsheets to complete the data migration.
Nocoly HAP can help clients develop business applications with low costs and high efficiency. The worksheets and relationships are implemented through the structuring introduced in this section, which is not complex.
Modularization
Similar to modularization in industries such as manufacturing and construction, modularization in digitalization refers to reusable business units or processes, and it emphasizes the standards which modules comply with so as to work together.
The essence of the Digital Middle Platform is business modularization. In large firms with complex organizational structures, there are likely to be multiple departments, each with partially overlapping functions. The key point of digitalization is to reduce or eliminate such overlaps, which is why modularization is necessary for large firms.
Here are three examples:
- All the departments are related to finance. Business activities with financial impacts can potentially affect a company’s income statement and balance sheet, and financial analysis is often concentrated at the management level. This necessitates that each department, when building their own applications, utilize unified financial dimensions (i.e., a unified financial account coding). In fact, both business and financial systems employ modularization, allowing for integration between them. We call this process the “Business-Finance Integration”.
- Sales, service, marketing, and other stages all generate and update customer information. Therefore, a modular system will establish customer data or enable customer data from one stage to be used by others. With this literacy, enterprises have the opportunity to view a customer’s full journey, from marketing touchpoints to after-sales service. We call this process the “Customer Data Platform”.
- Store brands need to manage multiple ecommerce platforms and their own stores, requiring a unified logistics and warehousing platform to provide accurate inventory information. This inventory information must be connected with procurement and production systems to ensure that the movement of every Stock Keeping Unit (SKU) is precisely recorded. Whenever enterprises add any procurement or sales processes, they write and read data to and from the logistics and warehousing system through a consistent interface. We call this system the “Modular Supply Chain System”.
The above demonstrates that modularization is the core of digitalization for enterprises.
Up to now, spreadsheets can no longer meet the digital needs of enterprises, while high-quality modular products have a significant competitive advantage and have become a business source for the software industry.
Automation
In enterprise management, there is a saying, “If machines can do it, then do not let humans be involved”, which reflects the concept of “automation”. Machines are more efficient and accurate than humans. Furthermore, the reason we invest effort in recording, classifying, and structuring data is to enable software and hardware to automate business processes.
For instance, upon contract signing, the delivery process can be automatically executed; after product delivery, financial processes can be automated; when a new employee joins, onboarding training can be automatically initiated; and based on the preventive maintenance schedule of production equipment, inspection tasks can be automatically generated. These examples all illustrate automation within enterprises.
To understand automation and design automated processes, it is essential to grasp the logic of automation. All automation consists of two elements: “triggers” and “actions”. Triggers are the conditions that initiate the execution of automation. It can be the addition or change of one record, or a specific time. For example, the change in the status of products to “delivered” can be a trigger, and the corresponding action can be issuing an invoice.
With the literacy of automation, we can identify the processes that should be automated, and determine the triggers as well as the actions.
Actually, the workflow of Nocoly HAP is designed in such a way that users can configure workflows visually without coding.
Reasoning
Apart from automation, structured data serves reasoning, which may be the start point of digitalization for the management personnel. If we want to know the impact of raw material price changes on profits, we need to structurally record costs for raw materials and compare it with profits. If we want to analyze the impact of equipment parameters on quality outcomes, we need to collect time-series data on processing steps and equipment parameters, and compare it with quality data that contains the same labeling information.
In fact, all reasoning aims at a particular instance of decision-making, which could be an operational improvement or a more strategic decision. Moreover, digital decision-making stands out as the pinnacle of digital literacy, and its prerequisite is high-quality data.
Reasoning can be divided into three parts:
- Setting Objectives
Setting objectives is a fundamental practice in modern enterprise management. Objectives should align with the SMART principles (Specific, Measurable, Attainable, Relevant, Time-based). Examples include sales conversion rates, defect rates, and employee turnover rates. Business analysis (Reasoning) should not only provide indicators but also automate calculations and even offer dashboards for users to gain a comprehensive understanding.
- Attribution and Correlation Analysis
Improvements should be based on data, rather than on the feelings of managers. For instance, quality assessments should be based on benchmarking analysis of specialized indicators against customer needs and industry standards. When operational indicators are low, data analysis should be conducted to identify the attribution or related indicators. Continuous improvement of related factors and verification of indicator changes can lead to better practices, which is the PDCA cycle.
- Decision Making Assistance
Information systems cannot replace enterprise decision-making, but they can assist in making decisions. Enterprise decisions include operational decisions and strategic decisions. Operational decisions rely more on the outcomes provided by information systems. For example, in procurement decisions in manufacturing, the quantity and timing of purchases depend on the structured data of production, sales, inventory, and material. However, strategic decisions are more sensitive to the external environment. Enterprises may not be able to obtain assistance from systems, so they rely more on the managers.
Our Practices
Through the aforementioned examples, the concept of digital literacy should be easier to understand.
Nocoly.com Limited, as a software company, focuses on these six aspects when training and assessing its employees. This is because professionals in this industry not only need to possess these digital habits but also have a toolkit to help clients achieve digitization.
Similarly, Nocoly’s product design and development also revolves around these six literacies. By Nocoly HAP, customers can build the basic elements of enterprise-level applications without programming. Data can be recorded, categorized, and structured through worksheet and view. By establishing relationships, data integration and modularization can be achieved, starting from one application to the entire operational process. Automation can be achieved with workflow, and data analysis can be conducted through a charting tool.
Once you’ve been using Nocoly HAP for a while, it becomes difficult to get rid of these digital literacies, and your team will be equipped with the best toolset to achieve their digitization goals.
