As technology evolves, IT is gradually facing more and more challenges, especially with regard to data governance. On the other hand, Jiuzhoutong Pharmaceutical Group is not afraid of difficulties and dangers in IT construction, independently develops ERP system and logistics system, and solves its own problems while innovating and investing in commercialization, providing services for the same industry and establishing a benchmark image.
The following is to share the data construction experience of Jointown Pharmaceutical Group . The original text is Liu Huanqing, Director of the Data Support Center of Jointown Operations and IT Management Headquarters, giving a speech at the FanRuan 2016 Pharmaceutical Conference.
About Kyushu
Jiuzhoutong Pharmaceutical Group Co., Ltd. is a large-scale enterprise group mainly engaged in western medicine, traditional Chinese medicine and equipment, with medical institutions, wholesale enterprises and retail pharmacies as its main customers, and providing customers with various value-added services such as information and logistics. . Based on the pharmaceutical and health industry, the company is one of the few companies with a nationwide network in China's pharmaceutical business field, and is in a leading position in China's pharmaceutical business industry.
From the perspective of technical support, the company has nearly 400 professional and technical personnel in ERP , logistics LMIS , e-commerce research and development, Kyushu Tongda Technology Development Co., Ltd., system implementation, and integration .
IT system architecture
Jointown's informatization platform pays attention to the principles of data standardization, resource flexibility, capability service, service networking, management automation, and centralization of management and control. The IT system construction aims at centralized deployment, reduces the complexity of system deployment and operation and maintenance, improves the interaction and collaboration efficiency between business and management units, and reduces the total cost. No system is deployed locally, only clients such as ERP are installed, and an independent LMIS system is not used. A set of ERP system independently developed by the group - JZTERP is used to manage business and logistics. The group headquarters has a unified platform. Logistics and wholesale are all sub-systems. The headquarters controls distribution and price strategies.
Data Integration Architecture Model
The figure below is the architecture model of Jointown Pharmaceutical Group, the core part of which is the data indicator database. The indicator database is a " language " for data exchange among various systems of the enterprise , which unifies the data caliber reported by various business departments and improves management efficiency. After opening up the indicator database, on this basis, Jointown has built a big data platform, data warehouse and business intelligence ( BI ) platform, and established various business management and control systems and medical data buses of the group . Self-service analysis , the processed data is displayed uniformly through the FineReport platform , and a unified data interface service is developed.
Master Data Governance Overall Solution
In 2008 , the group began to carry out master data planning, formulate management and standard-setting procedures, and clarify organizational division of labor. After that, the data is cleaned, the system is developed, and the implementation is launched. After 2009 , the developed system was promoted in the whole group, and the data gradually began to circulate within the group . After that, the second phase was mainly promoted in an all-round way.
In the second phase, the data service system DSS was launched . From the perspective of solving the data problems faced by the target, first of all, on the whole, it is necessary to build two types of databases at the logical level:
1. Unified database: solve data consistency, optimize integration system and other problems;
2. Private database: Solve the storage problem of private data of each application. Because it does not need to be shared, it exists as a private database.
The construction of master data enables the free flow of information and improves the efficiency of business realization; it provides valuable data -driven data for subsequent data analysis and scientific decision-making , and improves management level.
Business system deployment
JZTERP的部署方式有全国集中、区域集中和全分布式三种。基于中心性能、复杂度，以及网络风险等考虑，建设初期采取集团ERP中心+区域集中ERP部署策略。在网络风险降低的前提下（如4G商用），向大集中过渡。未来将来在条件成熟时，将区域集中的ERP向集团大集中迁移，中心ERP与各区域ERP结构一致，并且在ERP实施、运维过程中，始终保证全集团ERP版本的一致。
从集团决策、业务管控、系统运维等角度考虑，越来越多的国内外大型企业集团ERP部署已经或正在走向集中化。JZTERP采用集团管控与区域运营的合理兼容与整合，同时提供良好的伸缩性，为从分布式→区域集中→大集中的演变提供快速配置化支持。
- 集中的业务管理：实现价格、限销、资信、窜货等集中、分权管控；
- 集中的资料管理：简化各级公司管理工作，并防范质量风险；
- 集中的调拨管理：基于库存共享，集中调拨变为可能，并简化公司间销购单据转换；
- 集中的系统管理：降低分子公司技术要求，管理成果推广更加简便，并减低用户调动后的操作培训成本；
- 集中的审批流程：集成工作流实现上级公司集中审批下级公司的各项业务过程，比如退货、资信等；
- 集中的业务数据：数据集中合并分析及时、准确，为更多基于数据的集中应用提供良好支持。
- 本地化数据存取：保证分子公司业务操作高效率；
- 本地化业务操作：保证分子公司系统可用性，防范断网风险。
数据仓库建设
集团数据仓库成为三层，ODS层、DWH层、APP层。分公司的数据实时录入到集团系统，经过ETL清洗处理存储到数据仓库，利用FineReport进行前端数据展现。
数据仓库建设，讲究三个目标：1、集中：ERP、财务、人力、连锁、生产系统的数据都可集中到数据仓库。2、分离：数据仓库的作用既可做到数据存储，也可对其进行开发和业务分析。3、开放：所获业务数据可以再数据仓库基础上进行存取、应用和开发。基于业务系统——数据仓库——前端分析（FineReport）这样一条脉络，集团实现了统一化数据管理和分析。
- 为自助用户提供数据分析人员范围内数据自我获取、分析的功能。告别多个性化数据完全由信息部提供、使用人再加工的模式；提供安全、快速、及时、低成本的数据获取手段。开创集团数据使用2.0模式（自开发、利用用数据库引擎、大数据平台的自助、高效、及时分析能力）。
- 开发的数据展示平台提供了丰富的分析维度、提供信息的渐变性信息查询、全面梳理、支持日常考核及业务的管控、数据大集中的平台（批发、合资、中药生产、连锁、电商、人力、财务），协助企业管理层加强控制经营管理。
- 技术上实现各种数据的整合集中，对数据的综合性分析。为BI提供逐渐完善、干净、一致性的数据源。为领导层提供决策服务。
业务管控分析
业务分析决策基本上就是帆软报表制作的，企业需要做的就是平台支持，数据库，人员储备等措施。
绩效管理决策
BI决策管理针对的是企业高管，是数据化运营的核心部分，能对数据做到及时监控，综合反映企业运营状况。BI的建设在当前数据仓库汇集业务数据的基础上，将数据源扩展到财务、人力、物流等系统，综合反映数据，帮助决策分析。
BI效果展示：
未来规划
未来将进一步拓展大数据应用，集成大数据分析的业务决策。支持基于大数据驱动的精准营销，并以客户为中心，借助电子商务，移动商务等手段，建立端到端的客户服务流程。
技术上，采用业界通用的大数据系统和分析方法、模型，建设大数据平台；采用成熟组件进行低耦合的集成；以集中式部署降低建设成本和运维复杂度；抓住主要矛盾，循序渐进实施。
目标是要通过大数据深度分析挖掘，寻找更多的营销机会，让经营活动更具针对性，提升营利能力；通过大数据深度分析挖掘，优化库存商品结构与物理布局，提升物流作业效率；依靠大数据驱动，以及与各应用系统的集成，实现端对端的业务和服务流程。