Shulou(Shulou.com)12/24 Report--

Under the background of the accelerated integration of digital technology and traditional industries, various industries have moved towards the road of digital transformation, and the retail industry has always been at the forefront of the trend. As a legume family deeply cultivated in the field of plant milk for more than 80 years, the digital transformation of Vitasoy is also at the forefront of the industry.

At the recently concluded "Guanyuan data 2023 Intelligent decision Summit and Product launch" analysis practice sub-meeting, long Fei, director of data analysis of Vitasoy, gave a wonderful speech with the theme of "making data visible and palpable". Share the course, specific practice and three major strategies of Vitasoy's digital transformation and key data construction, so as to give full play to data value and stimulate business potential for more enterprises. Help the business to achieve leapfrog growth to provide value reference.

The following is a transcript of Longfei's speech:

The process of data Construction of Vitasoy: from the landing of BI products to systematic Management

In recent years, Vitasoy has started digital construction. Data construction is a very important part of digitization. In 2020, Vitasoy started the construction of BI platform, hoping to build a modern, high-performance, user-friendly and fast iterative BI platform. At that time, we took sales as the cut-in scene, mainly sales headquarters users, hoping to solve the problem of sales user data presentation and data sharing with the help of BI platform.

The year 2020 is the period for the introduction of the BI platform, and we have 100 + active users. By 2021, on the basis of deepening sales-related scenarios, we have also begun to expand our BI capabilities horizontally to finance, supply chain, manufacturing and other business areas. After more than a year of construction, Vitasoy has basically established a set of enterprise-level analysis system, the company's main business departments have become BI platform users, at this time the number of active users reached 1000 +. This year, we began to establish a relevant operating system, and gradually promote self-help analysis, through efforts, the number of active users has reached 2000 +.

In this process, we continue to receive business feedback, which is also a problem or pain point encountered in the data construction of many enterprises. Some of these problems can be collectively referred to as "invisible, invisible", such as no data support for decision-making, inconsistent data caliber between different departments, untimely data update, and so on.

Other problems that can be collectively referred to as "unnecessary, unusable and uncomfortable", such as data scattered in various business systems, need to rely on data analysts to manually export from different systems and then use Excel for processing, as well as the lack of data exploration tools, fast business changes and slow IT response.

How to solve the above pain points? Our understanding is to "make the data visible and tangible".

Make data visible and palpable: thinking and innovating from a business perspective

What is "making the data visible and palpable"? Let's take a look at it:

1. What is "visible"?

Our understanding is: let the correct data, in the appropriate situation, be used by users in need.

Correct data

Correct data not only refers to data quality, but more importantly, it is necessary to have a unified index caliber to get through the understanding of users at all levels. Also need to have the correct data granularity, to develop different data granularity based on the scenario to meet the needs of target users.

A suitable scene

One is the multi-channel access mode to achieve active and passive data acquisition. For example, Vitasoy BI has built comprehensive access to PC and mobile portals, and supports the push of email, enterprise micro, and BI embedded applications. The second is to fit the presentation form of the scene to help users understand the data quickly. For example, through a high summary of KPI Kanban to meet the high-level users' attention to the health of the whole business, and through two-dimensional tables to meet the needs of middle-level users for detailed data.

Users in need

With the right data and scenarios, how to get more people to use it quickly? Vitasoy has created a role-based permission package, and different roles have set a fixed permission package, which can quickly activate permissions according to the package to achieve efficient promotion while ensuring data security.

A specific practical example of "making the data visible" is "Manufacturing KPI Kanban": we have many factories across the country, each factory has multiple production lines, involving many core KPI, such as output, efficiency, quality and so on. The senior management of the manufacturing department hopes to know whether this part of the business is healthy through the data. We use KPI Kanban and mobile push to keep track of the health of the business in real time, and allow different functional departments of the factory to take relevant improvement measures according to the data early warning, and the progress of improvement will be further pushed to form a closed loop.

According to the latest platform statistics, the number of monthly active users has reached 2000 million, the number of monthly users is 60 million, and the user activity rate and report activity rate are more than 70%. BI has become an indispensable tool for the daily work of our business.

2. What is "touching"?

Our understanding is: to make themed data meet the needs of flexible users on easy-to-use tools.

Themed data

After the digital construction in recent years, Vitasoy has deposited a large amount of data, how to make it become "thematic data"? One is to analyze the theme, according to the analysis object to do topic classification, to meet different analysis scenarios. The other is hierarchical data modeling, which is layered according to data granularity, taking into account both efficiency and flexibility.

Easy-to-use tools

In the process of promoting self-help analysis, business users often worry about not being able to use it and not using it well, so we need lower barriers and easier-to-use tools to address business concerns. Also consider whether the tool can support advanced applications of other business users who can write SQL and python, so the tool should be able to get started quickly and have more functions.

Flexible demand

With themed data and easy-to-use tools, we can promote self-help analysis, let business users give full play to their initiative and creativity, and further unleash the value of the data.

A specific practical case of "letting the data touch" is the "big screen of financial data". It should be emphasized that this case is completed by business users with self-help analysis: financial users often need to pull all kinds of data when reporting, and then use Excel to make PPT into a big screen to report. We recommend the large data screen module of BI. Within a week, users used the cards and themed data sets in BI to build a large screen, won the praise of CFO during the report, and quickly spread throughout the financial department.

After the self-help analysis ability brought by "touch" has been promoted in the company, according to the latest platform statistics, 110 + users have participated in self-help analysis training and passed the exam certification. The number of monthly active users (Kanban or reports created in that month) is 80 +, with an activity rate of more than 70%.

Three key strategies of data construction: system construction, user identification and operation management

Based on the above practice, we summarize three key strategies for data construction, including:

1. System construction: a modern, high-performance, user-friendly and fast iterative BI system

The construction of the system should be based on business scenarios and requirements, including four key points:

Modern BI

Use data visualization to enhance quick insight into the business. In practice, we have found that a very important modern BI capability is the mobile BI application. About 60% of our users will access it through mobile phones and 40% through PC. Mobile office has become a major trend.

Efficient data model

Meet the needs of business fine management data. Including the above mentioned for different levels, provide different data granularity, as well as provide detailed data and so on.

Friendly transaction interface

Reduce the threshold of business use, improve the business experience of a more friendly interface and multi-channel access, can enable the business to achieve active and passive data acquisition, as well as scene-based interaction.

Fast iterative ability

Continuously improve the system capability to meet the flexible needs of users. In practice, we will collect product requirements and co-create data from different business projects, and also look for landing scenarios from new products, modules and functions of Guanyuan data. According to statistics, at present, we have upgraded 84 products, submitted demand for nearly 400, the use of Guanyuan data products is very in-depth, Guanyuan data also provides very strong support.

2. User identification: get to know users, relocate, and mine users' core needs.

The purpose of user identification is to really understand "user needs". We divide users into three categories:

High-level users

Mainly refers to the senior management of the enterprise, is the VIP user. The characteristic is that the expectation of data goes far beyond the direct supply of the system. Our strategy is to focus on solving the core problems, follow the "28 principles", and meet the business health analysis needs that high-level users like through mobile phone access and push methods that they are most concerned about.

Middle-level users

Mainly refers to the headquarters of the functional teams or regional management teams, is the key users. It is a connecting link between the preceding and the next, and it is very persistent in details, which is characterized by a lot of data requirements and flexibility. The strategy we adopt is to give priority to solving its "detail anxiety", and to give middle-level users the ability of self-help analysis with a themed hierarchical data model to meet the needs of flexible analysis.

Front-line executive user

This part of the largest number of users, but also need the simplest and most direct way to use data, is the main users. Our strategy is to address their daily work needs and give them clear and easy-to-understand data. Will also use the mobile phone to add push to meet their core KPI tracking and other needs.

3. Operation and management: systematization, data and service to promote the construction of BI

The systematization of BI construction includes operation and maintenance system and authority system, data includes BI platform monitoring, and service includes promotion and training, and business self-help ability, which are also the five key words of operation management.

Operation and maintenance system

One is report layering, which is hierarchical according to usage and user importance, and guides the priority of operation and maintenance through report layering. The second is to establish a hierarchical, sub-role problem handling process and related review mechanism to improve the efficiency of operation and maintenance and achieve closed loop.

Authority system

Distinguish report permissions from functional permissions, and divide role permissions according to the roles of different users. The self-help analysis users are divided into primary users and advanced users, and match different permissions at different stages. By improving the authority management system, taking into account both efficiency and safety.

BI platform monitoring

Establish BI platform monitoring screen, master the operation efficiency of the whole platform, timely take off the inefficient report, recover system resources, and improve the efficiency of operation and maintenance. Users' habits are found through platform monitoring to improve the efficiency of project design. At the same time, through the long-tail reports monitored by the platform, we can find opportunities for promotion and training, and improve the activity rate of users and reports.

Promotion and training

In the course of the project, three things should be done well for promotion and training: first, to invite high-level users to participate as far as possible in the project establishment phase, collect core needs, and do a good job of "pre-sales service"; second, take the initiative to provide predefined plans to users in the process of project implementation, and regularly report and synchronize, so as to be a good business partner; third, to do a good job of "after-sales service" by issuing announcements and pushing.

Training is inseparable from the course, and we have considered three types of course design: first, the guide for newcomers, including system operation, system function, process explanation, frequently asked questions, etc.; second, the basic and advanced application courses to help improve skills, and provide relevant examination certification. The third is the hot scene course, which makes the recent hot scenes of the company into courses and pushes them to users, such as how to use BI self-help analysis to achieve regional merger when the organizational structure changes.

The training forms include online courses that can make full use of fragmented time, online live broadcasts with wider coverage and lower requirements for venues, and offline courses that are more focused, efficient and targeted.

Business self-help ability

To improve the ability of business self-help, we should have both ammunition and management. First of all, through the unification of data portal, UI, layout specification and report naming standard, we can unify the user experience, let users form data culture and form a unified communication language. Secondly, provide complete user documents and update them continuously, including platform operation manual, online report list, report authorization baseline, frequently asked questions and orders, platform dynamic announcement, etc. Finally, a clear process guidance is provided to let users understand the BI permission application process, BI demand / service process and so on. Only when these businesses are done can we really use the various services we provide on the platform, thus releasing the energy of IT.

Finally, it is concluded that making the data visible and tangible is a necessary condition to realize the value of the data. It is "visible" to let the correct data be used by the users who need it in the right situation, and it is "palpable" to make the themed data meet the flexible requirements on easy-to-use tools. System construction, user identification and operation management are the key strategies to make the data visible and tangible. Data teams are not only producers and operators of data, but also do a good job in business empowerment and data operation.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.