Core content: supply chain management, amoeba management, four major business characteristics of the energy and chemical industry, six major management status quo, five major issues of management and operation data, four characteristics of energy and chemical industry data, Data decision management support solution (PC integration, mobile office, WeChat integration, large visual screen)
Under the dome: the background of the development of the energy and chemical industry
In June 2014, the "Energy Development Strategic Action Plan" (2014-2020)[1] was released, focusing on controlling total consumption, ensuring energy security, controlling coal consumption, optimizing energy reform, and promoting energy system reform. [2] In 2014, my country has become a net importer of coal, oil, and natural gas. The foreign dependence of oil is as high as 60%, and the dependence of natural gas exceeds 30%. The two together account for 15% of total energy consumption and imports. Energy consumption is increasing day by day, and energy security is a worry. [3] As practitioners in the energy and chemical industry, we should focus on the general trend of the world energy industry, pay attention to the latest technology, but start with the production and management of our own enterprises. We will use the most advanced technology and the most professional technology to promote the rapid, stable and long-term development of our enterprise. Accumulate a few steps and even a thousand miles, thick accumulation and thin hair.
- Strategic Significance of Data Analysis in Energy and Chemical Industry
At present, big data analysis has become a trend in various industries. In the use of big data, the energy industry also needs to work hard to catch up with the pace of society, and use massive data through specialized and specific analysis and processing for precise marketing, optimization of supply chain, quantitative internal management, optimization and monitoring of production. After years of construction of informatization systems, how energy and chemical enterprises can reasonably use data to obtain valuable information and provide decision support for the company's management is the main problem that needs to be solved in enterprise informatization. [4] The data analysis system is a management and decision support system based on relational database and multi-dimensional data warehouse, which can provide support for high-level decision-making of the company.
To solve the problems of data management and decision support in the energy and chemical industry, we must first start with the business characteristics of the energy and chemical industry.
- Four characteristics of business in the energy and chemical industry
The energy and chemical industry is resource-intensive, technology-intensive, equipment-intensive, personnel-intensive, and highly closed. my country currently has more than 30,000 energy and chemical enterprises, accounting for 73% of the total number of industrial enterprises. Among them, there are more than 10 super-large energy and chemical enterprise groups with total assets of several billion yuan, and medium and small energy and chemical enterprises account for more than 99% of the total number of chemical enterprises. These large, medium and small energy and chemical enterprises have the following four
Feature 1: Strong resource dependence. The resources on which the energy and chemical industries depend are mainly minerals, coal, oil and water. The per capita output and reserves of these four resources are very scarce in my country. The exploitation and utilization methods are extensive, the comprehensive utilization is low, and the waste is serious. Therefore, saving resources and improving resource utilization have become the primary problems to be solved.
Feature 2: Strong technology dependence. Accurate selection of chemical development fields and grasp of the future development trends of their technologies are crucial for chemical companies. However, at present, the new products and new technologies of my country's chemical enterprises are seriously insufficient, and the lack of technical level is the biggest bottleneck affecting the development of chemical enterprises.
Feature 3: The production process is special. The energy and chemical industry mixes or separates, extracts, and combines various components through energy, equipment and other resources, and causes chemical reactions, so each process can require the input of certain new components or resources (raw materials, catalysis, artificial, artificial, machinery, energy, etc.), and produce a number of outputs. The process is interlinked, the direction is irreversible, and the timing is strong.
Feature 4: Equipment specialization is strong. The storage equipment in the energy and chemical industry is mostly tanks, boxes, cabinets, barrels, etc., and most of the stored quantities can be measured by sensors. The production equipment is a fixed production line, maintenance is particularly important, and no faults can occur. When supply and demand change, production can only be maintained by adjusting process parameters without interruption.
- Six Status Quo of Energy and Chemical Enterprise Management
Because of the strong technical dependence of the business, the special production process, and the strong specialization of equipment, the management of the energy and chemical industry is more complex and interlinked, which can easily lead to loose management, inefficiency, and slow response and execution. But at the same time, due to the characteristics of strong resource dependence, efficient management is very helpful to save resources and improve resource utilization. Especially for a country with limited per capita resources, it is of great strategic significance to use advanced management methods (such as the current amoeba management) to improve resource utilization. So what are the six current status quo of energy and chemical enterprises?
Status 1: Business process management is chaotic. Data sharing is low and information acquisition is not timely. There may be a large number of independent excels for the statistical data of the production department, and multiple people manually reconcile each time.
Status 2: The degree of plan execution is low. The plans formulated at the beginning of the year and at the beginning of the month have no data support. The planned tasks are arranged subjectively, and there is no data monitoring in the whole process. The completion of the plan can only be found at the end, and the plan execution is difficult to achieve.
Status 3: Loose inventory management. Inventories are in various forms such as cans, boxes, cabinets, and barrels. Multiple units and personnel are responsible for inventory supply management, and it is difficult to share information with each other, highlighting the weakness of inventory management informatization.
Status 4: Cost accounting is complicated. In some processes in the chemical industry, multiple raw materials are input, and each process has a product output. In some processes, multiple products are simultaneously produced in one furnace, including joint products, by-products or intermediate products. How to allocate the cost to each product is a difficult problem faced by enterprise management.
Status 5: Inventory measurement is difficult. The chemical industry measurement is generally difficult to be very accurate, such as: the measurement of bulk raw materials is basically the weight of the goods minus the weight of the vehicle; the liquid or gas is mainly measured by pipelines, and the viscosity cannot be considered; the physical objects are mostly stored in the open air , the physical inventory can only be roughly estimated according to the volume. Therefore, how to solve the inventory measurement problem is another difficulty faced by chemical enterprise management.
Status 6: Slow response to data information. The daily business data management of chemical enterprises is extensive. From the headquarters to the branch companies, the decision-making information is not timely and accurate, resulting in the situation of "the generals do not know the soldiers, and the soldiers do not know the generals", which increases business risks.
- There are five major problems in the digitalization of management and operation
Chemical enterprises have a special business and have a large number of automation systems, which accumulate huge equipment data, production data, supply chain data and sales data. At the same time, due to loose management, the data connection of many systems of enterprises is not perfect, and some energy and chemical enterprises have a large amount of data, complex storage formats, scattered data, various types, and difficult applications. The information contained in different types of data also has its own characteristics. Only by integrating various data can it truly reflect the actual situation of the enterprise and be effectively applied to data analysis. It can be said that the informatization construction of chemical enterprises is facing huge challenges. [5]
Problem 1: There is a lack of uniform standards, and the statistical results are not uniform. The same materials and products have different names in different departments or different product lines. Finally, inventory statistics are carried out, and the results are various.
Problem 2: There is a lot of manual data, and there is a lack of unified integrated management standards. Different production systems, supply chain systems, supplier systems, sales systems, etc. all have the situation of manually exporting Excel, and production data statistics, a large number of Excel stubs. Excel itself is still in the form of various cross-grouping tables, and no system can do the support management of these daily updated data.
Problem 3: There are many manual statistics, many data reporting processes, and poor timeliness of aggregation. The business changes quickly. For the convenience of statistics, Excel is directly submitted. The workshop staff reports to the team leader, the team leader reports to the director, and the director reports to the leader. Each layer is a time delay and there is also the risk of data being tampered with because it does not meet the auditors' expectations.
Problem 4: There is a lack of efficient data analysis methods and means, and Excel is the golden key. The process and method of Excel analysis cannot be solidified, and when the amount of data is large, it cannot be satisfied. Every time data analysis needs are encountered, it is necessary to start processing and analysis from the original detailed data, and a large number of analysis methods that have been accumulated have not been precipitated into automated program analysis. It not only wastes labor, but also reduces efficiency.
Problem five: lack of overall management and operation data support structure planning, no effective data decision support system. The fact that there is no system that can do it automatically and accurately is the key to these five problems. Enterprises can get rid of the headache of "making reports every day, and they will be out of date" by precipitating on their own or learning from the experience of their peers' mature decision-making analysis systems.
- 能源化工行业数据四大特点
数据是管理出来的。能源化工企业普通存在的管理问题自然也影响到了管理数据化，不能高效的利用数据来支持管理和决策自然也就导致数据价值得不到重视。加之能源化工行业的业务特点，行业数据本身也呈现出不同的四大特点。
特点一：体量大。目前因为业务上技术依赖性强、生产流程特殊、设备专业化强，能源化工行业的管理显得更为复杂，环环相扣，很容易造成管理松散、低效，并且响应执行迟缓。但同时，因为资源依赖性强的特点，高效的管理十分有助于节约资源、提高资源利用率。尤其对于我们这个人均资源匮乏的国家，用管理手段提高资源利用便有着极高的战略意义。，中国能源化工企业100强的日数据生成量近一半都多于1GB，更有4.9%的企业 超过1TB。中国能源化工企业级数据中心数据存储量正在快速增长，非结构化数据呈指数倍增长，如果能有效的处理和分析，非结构数据中也富含了对企业非常有价值的信息。
特点二：种类多。采购系统、销售系统、仓库系统、客户关系管理、生产管理系统（主生产计划、物料需求计划MRP等）、财务管理系统、视频监控系统、GPS物流管理系统等各类数据，有日志数据、文本、图像、视频、音频、关系型数据库、多维数据库等等。
特点三：价值密度低。能源化工行业自动化系统多，每一秒都有大量的系统监控数据和业务数据自动生成，但实际上能用来指导决策分析的并不多。大量的数据需要做基础汇总之后再做分析。
特点四：速度快。能源化工行业自动化系统24小时不停线运行，成集群模块化分布。数据传输速度快，这就要求数据决策系统要能及时响应高速更新的数据，未业务运营提供实时的分析汇报。
数据决策管理支持方案（销售、供应链、财务）
面对能源化工行业管理经营数据化的五大问题和行业数据的四大特点。传统的解决方案显然难以满足多层次的业务需求。以十几个能源化工数据决策分析系统的案例来看，分五层的规划设计（如下图）是目前能照顾到各层次需求的绝佳解决方案。今天笔者在这里着重聊聊数据展现层和数据应用层，前三层可以在今后和大家详细探讨。
- 数据治理和整合
前三层（数据源、数据处理层、数据存存储）主要内容就是数据治理和整合笔者简单介绍下。前文已经和大家聊过了，企业中有着大量的业务系统，如SAP、ERP、OA、CRM、SRM、EHR、MES、PIM等，同时还有大量的Excel手工帐。系统化的解决这些大量的不规范数据的办法就是构建企业级数据仓库（量大，低频更新）+ODS缓存区（量小，高频实时更新）。当然也存在直连业务系统数据库的方案，但直连的方案在性能上容易遇到瓶颈，同时对实际业务系统造成压力，大多数企业考虑到数据安全和业务风险，选用直连方案都十分谨慎，采纳实施的案例不多。
有了企业级的数据仓库，分析页面数据直接来自被打通的各业务系统。向上，可以通过全区域、全产品线、全业务、全系统的数据，汇总分析支撑战略规划；向下，通过对各地区分公司每日各项指标的把控和指导，把控业务运营管理的全过程，不同层级的职员，通过对汇总数据的细化，逐级分配到自己所负责的业务范围和人员，实现运营管理的数字分析决策。下图是向上涵盖战略和经营，向下涵盖管理和操作查血的具体展现层分析模块。
（笔者在此以销售与分销分析管理、采购与库存分析管理、财务分析管理来探讨具体的方案建设。考虑到数据分析模块的分析深度、文章篇幅和笔者精力，生产与技术管理、人力管理、市场客户管理、供应链管理、产品研发管理、经营仪表盘和指标库、市场监测、行业对标、战略地图等等模块的分析和探讨，笔者暂定在今后的文章中一一道来，也欢迎读者文末留言交流。）
- 销售管理的核心是订单过程管理
我们的一线业务人员通过查询每天的订单明细，实时掌握责任田的订单详细情况，及时掌握订单执行的详细节点。让业务人员能及时根据订单动态调整自己的工作重心。就好比我女孩子盯着天猫、聚美优品的订单物流信息，根据实时的物流状态，安排自己是否要今天出门。以此类似，业务人员也可以根据订单明细和动态，去安排自己是否要再次电话、拜访等跟进客户。实时的数据查询系统满足了一线业务人员提高效率、及时安排行程的核心需求，特别是移动端实时查询。
而作为销售部门经理，则更为关注短期内（一周/一个月）内商品价格走势、销量走势、人员的业绩走势、客户订单采购走势等。针对市场商品价格走势（虽然有些是协议价格，但不少业务产品仍然是随着市场波动，存在价格不稳定情况），及时调整产品价格；根据销量走势，及时关注库存和相应的产品，决定是否加大该产品相关的销售成本投入；根据人员业绩走势，及时发现业绩占前10%和后10%的销售人员，对优秀的销售经验及时总结分享，提高团队整体销售业绩，并通过即时奖励，提高团队士气和战斗力，而对业绩不佳的销售人员，及时发现问题，介入指导，督促改进，确保“伤员及时恢复战斗力”，如果最终通过几次数据考评，定性为害群之马，及时剔除，防患于未然；而对于客户订单走势，根据不同地区、不同行业、具体不同客户采购的订单量走势，及时进行客户关怀，把80%的精力用在20%的优质客户上，同时用20%的精力用在80%的普通客户和潜在客户挖掘上。数据分析系统通过实时的数据汇总和关联分析满足了销售部门经理及时掌握销售动态、调整销售团队管理、制定针对性的销售策略的数据支撑需求。
作为公司领导层，只需要关注几个固定的页面，就可以从宏观掌控销售订单的区域分析。还有什么是比这更能解放领导时间的呢？彼得德鲁克告诉我们，领导者的时间是宝贵的，那么节约了领导在具体事务上占用的时间，留下来更多的时间思考公司的发展和行业变化，这不也正是数据化决策所要实现的目标之一吗！
- 供应链管理目标是在现实的资源（资金、仓库面积、供应者）约束下满足订货的需要又能使成本达到最低。
圈子内的人，我们常说“库存是万恶之源”，减少不必要的库存，追求"零库存"成了我们日思夜想的事儿。正如李叔同的《晚睛集》里的词：“念念不忘，必有回响”（电影《一代宗师》中的赵本山饰演的丁连山也说过，估计大家更熟悉），我们实践总结出补给策略方法论:定量订货，定期订货。关键控制点分ABC分类控制，核心内容就是“关键的少数和次要的多数”。具体哪些关键的少数，哪些是次要的多数呢？未应用大数据解决方案之前，这个是通过业务经验资深的专家凭借多年实践积累摸索总结出来。时至今日，经验已经有部分过时，并且企业发展速度越来越快，已经等不得耗费多年的经验总结了。在此背景下，自动化的大数据分析方案和系统应运而生。
我们可以分析下图方案，通过系统的日均完成率对比，及时掌握了作为整体的一部分的零件的配比情况，及时对配件库存进行调整，保证整体产品的完整输出。实际产量结构分布图，则清晰、准确的看到日均计划和实际完成的差异，及时调整当日、次日生产计划；而正品入库产量完成，可以说是库存管理者交的一份答卷。之前的工作都是在答题，而这个正品入库产量完成，则直接响应了业务部门的销售提货需求。而我们最容易看得见的，就是这个正品入库产量的积压和不足。但实质的解决办法却是要监控调整从日均完成率对比和实际产量结构分布。一套自动化的数据监控与分析系统，让自动化的生产线装了一个决策大脑。而浙江传化集团的成功经验是实行供应链分析管理，2016年节约成本13%！
当然，这边也有另一种日库存分析的方式。如下图，同样是库存分析，区别在于无需多考虑库存对销量的影响，只是统计总数就可以了。这是较为粗犷但仍然高效的统计分析方案了。
下图中的指标：库存数量、库存金额，维度：产品(成品油第二层级物料，化学品第三层级物料)，通过查询和钻取，特定展示运营层领导比较关注的特定时间的物料库存情况，为运营策略的调整提供决策数据依据。同时也方便数据追踪和领导莅临审查。用这种自动化的报表分析页面去代替传统手工Excel，实现了历史数据可追溯、操作实时便捷的目标。
- 财务指标分析是指总结和评价企业财务状况与经营成果的分析指标，包括盈利能力指标、偿债能力指标、运营能力指标和发展能力指标。
财务指标是管理层和领导层同时关注的核心指标之一，财务指标分析是企业进行管理和改革有效性的第一衡量指标。财务管理的目标是实现产值最大化、利润最大化、股东财富最大化、企业价值最大化、相关方利益最大化。而信息化程度、财务架构是否健全、内控体系完善性、成本核算精细程度、费用管理规范性等都会大大影响财务的管理能力。
如下图分析展示，通过切换公司、产品和时间区间三维度，既可以从公司维度反应不同指标的获利能力，为公司绩效提供数据支撑，又可以从部门维度反应不同指标的获利能力，为部门绩效提供数据支持。通过这样的自主定制的可视化界面，解决阿米巴经营管理最难的成本核算和内部定价问题。A车间部门满负荷生产一批订单，需要暂时未满负荷运营的B车间部门协助生产以早日完成订单生产，B车间消耗了不同原料（比如煤炭和硫磺）为A车间生产了一定量的产品，最后A车间又把B车间剩余的一部分原料（煤炭和硫磺）运走了。这种情况，如何裁定A、B车间的成本和利润呢？通过下面这种灵活自由定制的页面，财务部门审核定价，实现各部门的成本独立核算，效益独立核算。从而避免了大量的手工Excel，同时高效的实现了精细化管理，让每个车间和部门都可量化投入产出。通过实行阿米巴管理，2016年营业额1个亿的恒逸石化，光采购成本节约就达到了17%。
通过切换不同的排序方式，我们发现销售公司和广东公司两个公司的获利能力是其他公司的两倍以上，那么是不是要有两点反应？第一点，这是集团公司的拳头业务子公司，重点要保持持续获利，应该挖掘出来成功经验，向其他子公司推广可复制的经验；而是，过年了，是不是该发年终奖了啊，给谁发，发多少呢？嘿嘿。
接下来我们看看不同公司的营运能力如何。通过历年分析，我们看到企业的总资产和净资产基、应收和现金流本保持逐年上升，当前年度2016年有所下降，那么具体是哪些部门哪些业务什么原因导致的下降呢，我们通过多层钻取去查找原因。我们总能通过公司、部门、产品线、区域、月份等维度钻取到是某些特定维度（比如说是区域）下资产和营收指标起伏较大。通过对比，把问题定位到具体的维度上（区域维度），同时再通过该维度（区域维度）的钻取，找到其他维度（比如说月份）指标起伏较大，以此逐层钻取，发现问题。
- 移动办公潮流，大屏可视化驾驶舱
随着互联网时代的飞速发展，手机正在全民中迅速普及。据台湾《电子时报》报道，工信部的统计数据显示，截止到2016年5月底中国的手机用户数量已达到12.56亿人，相较4月份增长了0.36%，比去年同期增长了7.82%，相当于中国90.8%的人都在使用手机。，在所有使用手机的人中，使用3G网络的用户有4.64亿人（占比36.94%），所有使用手机上网的用户数量为8.57亿人，占总数量的68.24%。以上数据表明，移动生活大潮已经来临，而移动办公潮流正在兴起。移动互联网时代，信息无处不在。充分利用移动应用，人们可以摆脱办公场所的限制，充分利用碎片时间，进而可以“管理于拇指之间，决策与千里之外”。笔者这里分享几个移动端效果、PC集成效果、微信集成效果，以及大屏可视化驾驶舱效果，给各位读者养眼之用。
如何看待数据决策
数据决策本身不是万能的，也不是凌驾于业务系统之上的。他是企业信息化发展到较高层次的产物。整个信息化发展可以分三层概括：运营层、管控层、分析层。其中企业中SAP、ERP、OA、CRM、SRM、EHR、MES、PIMS等大多数信息化系统其实都是解决了运营层的采购管理、供应链管理、研发管理、生产管理、库存管理、销售管理、客户管理流程、流程管理、人资管理等管理运营问题。而通过这些生产和业务相关软件内置的报表和流程功能，辅助财务软件（比如用友NC）实现了管控层的人、财管控。而在此基础上的分析层则更多依靠Excel类数据汇总、PPT制作报告、外聘行业专家提供报告等来满足需求。对于经营决策和经营会议，数据支撑起来的决策让领导层从直觉、感觉、经验逐渐过渡到逻辑、关联上来。能通过数据本身的变化和调整，直接在宏观上调整实际的业务经营。有了数据支撑的经营决策，外加不同的主题分析，像核磁共振一样对企业进行精准的全方位扫描和监测。如此，在战略决策的大方向上，有数据支撑作为依据，再辅助对历史规律、行业动态的把握，让决策更具可操作性。
最后
实现大数据分析价值的三大要素是支持、信任和技术。应用大数据分析的企业需要管理层持续的支持，需要加强跨专业部门之间的信任，并具有深层次的业务知识和技能。于此同时，大数据决策分析正方兴未艾，需要我们抱着探索的心态，勇于在具体的业务中亲自实践。
文中部分截图来自以下企业的项目实施方案：浙江海利得新材料股份有限公司、浙江传化集团、云天化集团有限责任公司、浙江恒逸集团有限公司、中策橡胶集团有限公司、中国海洋石油销售公司、旭阳控股有限公司等。
PS：文以载道，学而进阶，欢迎留言探讨
本文出自帆软数据应用研究院
— 帆软数据应用研究院 —
帆软数据应用研究院专注于企业的数据应用研究，致力于让数据成为生产力。主要分享行业趋势、市场动态、理论观点以及企业的数据应用实践案例。
参考资料：
[1] "Energy Development Strategic Action Plan" (2014-2020), Xinhuanet, [Published on 2014-11-20] The General Office of the State Council issued the "Energy Development Strategic Action Plan (2014-2020)" - Xinhuanet
[2] Relevant person in charge of Energy Bureau answered reporter's question on the "Energy Development Strategic Action Plan", the website of the Central People's Government of the People's Republic of China, [published on 2014-12-15] Relevant person in charge of Energy Administration responded to the "Energy Development Strategic Action Plan" Answers to reporters' questions_Department News_News_China Government Network
[3] The State Council issued the Energy Development Strategic Action Plan, Xinhuanet, [Published on 2016-11-20], the State Council issued the Energy Development Strategic Action Plan (2014-2020)
[4] Xu Bin, Wang Xiaodong, Lin Li, "The Way of Transformation, Upgrading and Competitiveness Reshaping of Big Data Management Enterprises". [M]. Beijing: People's Posts and Telecommunications Press, 2016.1:149,162
[5] Cases of the chemical industry - problems faced by the chemical industry, FanRuan official website, [citation date, 2017-2-6], Chemical Industry Information Construction | Chemical Industry BI Solutions - FanRuan Data
[6] Zhu Chao, Director of Chemical Industry of Fanruan Company, "The Way of Data Application in Petrochemical Industry", [R] Fanruan Company, 2016