Making a Shift from Descriptive to Prescriptive Analytics and Planning

Share on

FMCG

Making a Shift from Descriptive to Prescriptive Analytics and Planning

“Supply chains are undergoing rapid transformation with the use of machine learning and artificial intelligence, which can enable us to optimize our processes, reduce costs, and improve customer satisfaction. AI and ML based systems are the future of supply chains; however, I believe that the more pressing challenge is how to address the issues of data quality, skill development, and the transition from descriptive to prescriptive analytics and planning. We need to make sure the data we feed into our new systems is accurate and reliable. Moreover, we must train and empower our people to use the new technologies and methods,” stresses Saurabh Lal, Sr Director - Network Design, Logistics, Planning & Digital, Kellanova (Kellogg Asia Pacific Pte Ltd), during this interview…

How has the supply chain segment transformed over the years? What have been some of those defining moments?

Saurabh Lal

Supply chains have transformed over the years to cope with the changing business environment and customer demands. Some of the key transformations witnessed are:

  • Global sourcing and outsourcing: Many companies started to buy & sell goods & services from different countries, taking advantage of lower costs, similar quality, and greater availability. An interesting book on this subject is ‘The World is Flat’ by Thomas Friedman, which I read many years back. It analysed the impact of globalization and the opportunities it opened for people and companies worldwide. However, as we see today, globalization came with its own risks, which are playing out in the geopolitical arena today.
  • Manufacturing technologies: New technologies became available, reducing capital costs and allowing companies to make their products faster, better, and cheaper. I still remember in my first job; we had a person standing with a stop watch to measure machine efficiency and a single head form fill & seal machine controlled by cams was state-of-the-art. Today smart factories have allowed us to control processes to improve productivity and single head machines have given way to 10 or even 20-head fillers, which has dramatically improved outputs.
  • Digital platforms: Many companies use digital platforms – from the basic ERP transaction systems to advanced computing, data management and internet of things. These platforms enable real-time data collection and analysis, enhanced visibility, and collaboration. The days of worksheets made using ‘123’ and limited email availability are long gone.

You have witnessed the growth of supply chain vertical from close quarters, both in Indian and overseas. What are the factors that still demand attention from stakeholders in India to be at par with the rest of the world?

India holds immense potential in becoming a global leader in supply chain management, as it has a large and growing market, a diverse and skilled workforce, and a strategic location. However, it also faces some challenges and gaps that need to be addressed by the stakeholders, such as:

INFRASTRUCTURE: India is investing heavily in developing its road, rail, air, and port network, but it still lags other countries in terms of efficiency, and capacity. We need to improve warehousing and handling facilities, especially for small and medium enterprises, and prepare for automation, such as using palletized loads, material handling equipment, thus reducing manual work.

DISTRIBUTION NETWORKIndia's intricate distribution network poses a challenge for supply chain management, especially with diverse customer types like supermarket chains, kirana stores(small mom and pop shops), and e-commerce platforms. This complexity demands efficient logistics management and technology integration to streamline supply chain processes. The introduction of the Goods and Services Tax (GST) and infrastructure enhancements necessitate a reimagining of the digital supply chain to enhance cost-effectiveness, service levels, and customer satisfaction.

SUSTAINABILITY: India faces a major challenge in achieving supply chain sustainability. Reliable water management, and waste disposal, are still a challenge for industries though they are improving exponentially. Additionally, industries must comply with new regulations and standards, which are the need of the hour, such as the National Green Tribunal (NGT) and the Extended Producer Responsibility (EPR), that aim to reduce the environmental and social impact of supply chain activities. We need to foster collaboration among companies, government, and civil society.

SKILLS: India has a high literacy rate and a strong aptitude for science and technology. However, we also suffer from a demand-skill gap, with lack of practical and vocational skills that are required for supply chain management. We need more investments in skill development and training programs, both at the academic and industry level.

How can companies achieve supply chain network optimization?

Supply chain network optimization is the process of finding the best way to organize and manage your supply chain, such as where to locate your factories and warehouses, and the most efficient transportation modes. Here are some steps you can follow to optimize your supply chain network:

  • Decide what is your end goal and time horizon: The first step is deciding what you want to achieve with your supply chain network optimization, and how much time into the future that you want to consider. Are you solving for cost or service or are both to be balanced?
  • Map your current network and costs: The next step is to understand how your supply chain network works now, and how much it costs. Collect data on the costs and performance of each node, such as inventory, transportation, labour, and overheads.
  • Build a working hypothesis: Brainstorm possible solutions and develop some ideas on what you believe the future network may look like. Collect external data from similar industries as well as understand the drivers of your current network design. A new or modified network design may require adding, removing, or relocating some facilities, changing some transportation modes or routes, or adjusting some inventory or production policies.
  • Analyze parts of the network: Examine different parts of your supply chain network and see how they affect your goal and time horizon. You can use tools and methods such as SWOT analysis, benchmarking, or gap analysis, to identify the strengths, weaknesses, opportunities, and threats of each part, and compare them with the best practices or standards in your industry or market.
  • Test with simulations: You can use computer models or software to simulate your supply chain network and see how it performs under different scenarios and conditions, such as demand changes, supply disruptions, or price fluctuations. You will need a financial analyst to measure and evaluate the results, such as cost-benefit analysis, return on investment, or key performance indicators.
  • Ensure you manage risks and sensitivities: The final step is to make sure that your optimized supply chain network is robust and flexible and can handle any changes or uncertainties that may happen in the future. Risk analysis, sensitivity analysis, or scenario planning can help assess the potential risks and impacts of your supply chain network. Alternative actions can then be presented.

Keep in mind how will the change impact your products or your customer? Ensure that you take a business view in your network design and not just a supply chain view.

In one of your recent posts, you had mentioned, ‘Forecasts are always wrong!’. What are the driving principles of forecasting and how can companies address forecasting challenges?

Forecasts are predictions of what will happen in the future, based on past data and current trends. They can help companies plan their supply and demand and make better decisions. However, they are statistical ‘predictions’ and inaccuracy is inherent in the numbers. Therefore, companies need to use forecasts judiciously, and follow some principles, such as:

Choose the right timeframe: Forecasts can be made for different time periods, such as short-term, medium-term, or long-term. Companies need to choose the right time frame for their forecasts, depending on their purpose and goal. Long-term forecasts can be used for financial planning and capacity planning. Short-term forecasts can be used for fulfilment execution planning and daily operations.

Use probabilistic forecasts: Forecasts should not be a single consensus number which carry the flaw of deviations introduced during the IBP process. On the other hand, probability forecasts, which show the range of possible outcomes and their likelihood, can be more helpful to drive business plans while preparing for multiple scenarios or outcomes and built flexible response plans.

Decide on the level of detail: Forecasts can be made for different levels of detail, such as product, category, or SKU. SKU-level forecasts can be more accurate for short term forecasts, as they capture the variations and trends of each product. However, SKU-level forecasts will be significantly inaccurate in the long term, as they are affected by morefactors and uncertainties. Use higher level forecasts, such as category or product forecasts, for long-term forecasts, as they are more stable.

Update and adjust forecasts regularly but at an appropriate frequency at which your system can respond. It is not practical to forecast daily if your response time is a week. This will only introduce noise into the system. Information and assumptions can change over time, due to new data, events, or feedback. These can be used to develop latest estimates to adjust supply execution. It would be even better if a Demand Drive MRP or pull system can be developed to enable continuous replenishment.

What are the upcoming digital supply chain innovations and the poised role of AI in planning solutions?

Supply chains are undergoing rapid transformation with the use of technology in supply chain management, including machine learning and artificial intelligence, which can enable us to optimize our processes, reduce costs, and improve customer satisfaction. AI and ML-based systems are the future of supply chains; however, I believe that the more pressing challenge is how to address the issues of data quality, skill development, and the transition from descriptive to prescriptive analytics and planning. We need to make sure the data we feed into our new systems is accurate and reliable. This can be a major challenge for large or complex enterprises with multiple sources of data. Further, we need to train and empower our people to use the new technologies and methods. Weaning away from spreadsheets and presentations is required but perhaps not so easy.

What’s your take on sustainability in supply chain? How can companies manage Scope 3 emissions?

Companies have taken ambitious targets on the UN Sustainable Development Goals and Net Zero. However, achieving these is a complex and challenging task that requires data, connectivity, and collaboration among multiple players in the chain.

We must focus on completing scope 1 & 2 while progressing to scope 3. Scope 1 emissions are relatively easier to manage as they remain mostly in our control and various strategies are employed to achieve the targets. Scope 2 emissions or indirect emissions require generating companies, among others, to share data and collaborative work among partners. Scope 3 emissions are other indirect emissions that occur in the value chain of an organization, both upstream and downstream, such as emissions from raw material extraction, transportation, waste disposal, or use of finished products. These emissions are more difficult to manage, as they involve multiple suppliers, customers, and partners that may have different levels of maturity, transparency, and ambition in terms of sustainability. However, these emissions also represent a significant opportunity for improvement, as they often account for the majority of an organization’s greenhouse gas emissions.

There is a need for aggregators and climate platforms, which measure, report and reduce emissions for enterprises. We should see significant opportunity and execution in this space in the next few years.

What are the skillsets that new age supply chain professionals possess to tide over tough times or drive the next wave of supply chains globally?

Supply chain will require significant investment in human resources over the coming years as they continuously evolve to respond to market changes. Some key factors for new professionals will be:

  • Strategic thinking: Understanding how different parts of the supply chain interact and affect each other and designing and implementing plans not just tactical but also strategic in nature.
  • Analytical and data management skills: Collecting, processing, and interpreting data from various sources and systems, and using it to make informed and effective decisions and actions.
  • Adaptability and outward focus: Ability to cope with change and uncertainty, and to adjust to new situations and demands. It also involves being open-minded and willing to learn from external sources and other companies and success stories.

More on FMCG