Let us begin with a thought-provoking activity. Try to find out any three key data points about Bangladesh that you can rely on, refer to, and use to make a day-to-day personal or professional decision. It might be the population of the country, GDP, revenue, no. of students who will be sitting for HSC exams, actual generated power, etc. You might get some information through Google search or from related websites, but can you be sure about their credibility?
Let’s see another example. To learn about a company’s recent financial condition, let’s say General Electronics, this writer checks the data available in the Wall Street Journal or Bloomberg, and like an auto-reflex, his mind concludes about the company’s current condition without any confusion or skepticism.
However, data from the Dhaka Stock Exchange website regarding any local organization’s current situation does not assure him. Why is that?
Some key aspects of data are: a. What should be recorded? b. How should it be extracted? c. Where to be linked? d. How long will it remain valid?
To illustrate the critical importance of the above four factors, we may just recall the incident of the ‘Great Chinese Famine’, which lasted from 1959 to 1961, and more than 30 million people starved to death. The key contributor to this man-made disaster was a complete failure in data management (intentional or unintentional, that is a different angle). We can assess the four aspects of data in light of this historic disaster.
What is required to become a data-centric mind? How do you become someone who will rely on data rather than educated guesses (it’s just a sugarcoated term), hunches, or gut feelings? It seems losing weight and living a healthy life might be easier than living a data-centric life.
Most of us don’t even consider making an excel sheet or keeping a diary of daily life expenditures. We barely have any idea how much we spent during our educational life. That means we only make monthly plans which are also heavily dependent on lump-sum budgets.
If you notice carefully, you’ll be amazed to discover your day-to-day forecasting ability. From his experience meeting some European middle- and upper-class families during his graduate studies, he can confirm that they are opposite, meaning they need data to predict. This implies our inherent tendency to go with the flow rather than making the flow according to necessity.
Now, let us think about an industrial project or a manufacturing operation and how we manage data there. In layman’s terms, it’s a nightmare. Multinational projects or manufacturing organizations are bound to use their legacy ERP (enterprise resource planning) or MIS (management information system) solutions; thus, the people from those organizations in Bangladesh might not feel the same way.
However, the economy is heavily dependent on local organizations. For over 10 to 15 years, companies have started procuring or implementing ERP solutions in Bangladesh. But, only a few have successfully used such data management systems.
One key reason for that is a lack of proper transition plans. There are organizations where disintegrated paper-based log sheets were the sole operational data management tool. Without getting any hands-on experience with basic data management tools like Microsoft Excel or Access, some industries jumped to ERP-based solutions. The outcome was the failure of a system that has been globally successful since the 1990s.
The second major reason is the lack of proper use of instrumentation and electronic measuring systems in processes, projects, and manufacturing units. This is still a great problem in most factories, including some very large manufacturing facilities. Both reasons are due to our callous attitude towards the four key aspects of data management.
Still, there are major manufacturing conglomerates that tend to depend on experienced technicians or foremen who try to complete troubleshooting tasks by using audible sounds or visual aspects of a running machine rather than buying proper measuring tools and engaging proper engineers to diagnose the situation. This depicts the tendency of negligence toward data-driven operations. This attitude prevents organizations from being ready for Industry 4.0 and makes them blame ERP or any other data management system concerning their investments.
The world is moving ahead. Now, it is Industry 4.0, while AI has also arrived. Yet, we still do not have proper arrangements to feed timely and appropriate data automatically into ERP systems. There are organizations that have invested lots of money in systems like SAP and Oracle, along with yearly renewals, but are barely able to use them or, in some cases, have completely sidelined such systems.
Once again, the writer would like to share his experiences in heavy industries like power, steel, float glass, and chemicals and how they hurt our economic or industrial cycles.
The devil is in the inventory. In our industries, we lack a proper information management system for measuring the size of the inventory we are handling in terms of money and how much cash is blocked as inventory holding costs.
In one of my experiences, it was discovered that around BDT 4 billion was stacked as unused raw materials for two years in one giant heavy industry concern. The organization is a profit-making one. Now imagine what would have been the achievable profit margin if they could have managed it properly and not stacked it at the back of their raw material storage area. Adding to the disappointment, the organization used one of the most popular ERP systems to manage its operation.
Next in line is the operational availability of a plant or manufacturing unit. In our country, most manufacturing units do not have proper forecasting data analysis for predictive and preventive maintenance. This translates to an unknown number for a plant’s probable productivity.
Similarly, we do not have proper data about sales lead time or customer onboarding time. We do not know how long our sales forces are engaged for a particular client or customer and what the contribution to the balance sheet is. A similar gap can be found in the purchasing cycles. Tracking the international rates and tagging them with the purchases shows how efficiently and timely the sourcing work is getting done.
Now comes the most important portion – how to implement a data management or data-centric business operation for an organization. First, we need to set the owners of the critical data points. They will be the keeper of ‘Master Data.’ Master Data is the foundation of an operational cycle. For example, in a powdered milk manufacturing facility, the premium product grade can be master data. Say ‘PM 01′, and to label a product PM 01, some fixed parameters must be stored and tagged with PM 01 as the recipe. Similarly, a customer code or account can be master data.
Once the master data owners are fixed (obviously department-wise production, maintenance, accounts, sales, SCM, etc.), a coordinating data management team will be formulated. A 2- to 4-member team for a 5- to 7 thousand-employee-based organization should be sufficient. This team coordinates the bridge between the data owners and keepers of the day-to-day operational data points through existing systems, newly adopted systems, or basic Microsoft Access.
A top-down approach must be used to create a data-centric mindset in an organization. It will be a painful journey before it becomes the norm. This coordinating team should be answerable to and driven by the board, stakeholders, or company owner.
Now, from the organization’s top management, 5 or a maximum of 6 data points need to be identified as the organization’s KPIs. The data management team, along with the help of Master Data Owners, will create interlinking data sets and fixed formulas to get these KPI values in real-time or through everyday operations. So, it means the data trees for driving the organization have been penned down.
In the next step, there will be two routes. If the tasks mentioned above are completed through an existing ERP or similar solution, then there will be the journey of automating the data retrieval process through different business and operation cycles and cross-functional training.
For instance, IT/ERP teams’ training on basic operation and business cycles to understand the frequency of generating the Master Data and maintenance teams to be trained about basic database or ERP solution functionality.
However, if the primary steps are completed through some basic Microsoft Excel and network-based Microsoft Access solution, then a capable IT team with ERP solution expertise needs to be onboarded because there will be major tasks for a smooth transition to becoming a data-centric company.
If we divide the transition into phases, the toughest one will be the phase 1. Proper understanding of the board/owner/stakeholder and the mid-level management (master data owner) is the most crucial factor for success. Once this first step is implemented, the organization can visualize and create future road maps for the next steps, making the journey achievable.
Data-centric operation does not mean changing the organization’s culture. It means automating the culture so that the success that came through can be nurtured and sustained. Technology is nothing but a tool, an enabler. Once the culture is linked with a data-centric mindset, technology will bring success and contribute to the balance sheet.
In the end, we all want to be members of a manufacturing plant, a financial institute, or a society where most things will make sense and be logical. Reliable and readily available data is the foundation for such a system or society.
Giasuddin M Tauseef is currently the Country Manager of Bangladesh SMS Group. He is a techno-commercial project professional with an electrical power engineering background. He has 15 years of experience in heavy industries, especially power generation, distribution, steel, and non-ferrous manufacturing.
