Technology will truly start supporting supply chain planners has been saved
Technology will truly start supporting supply chain planners
The future of planning, part II
In 10 years’ time, planning will be a whole different ballgame. Now is the time to start preparing – in terms of technology, talent, and organization. In a series of articles, we will share our views on the future of planning. In the current article we will discuss how technology is taking over the majority of planning activities, leaving room for planners to focus on the exceptions.
Go directly to
- Technology is finally fulfilling its promise
- The innovations
- The future of planning series
- More information
Technology is finally fulfilling its promise
If there is one thing we can learn from unforeseen events like trade wars, political turmoil, earthquakes and, recently, the COVID-19 pandemic, it’s that markets can change overnight and we need to be able to respond swiftly. Fortunately, we are experiencing an exponential growth in digital capabilities. Technological innovations like machine learning are finally starting to live up to the 10-year promise of largely automated processes and exception-based planning. These innovations will hugely enhance (real-time) insight and support faster and better decision making.
At last: exception-based planning
The breakthrough of exception-based planning is based on growing computational power to handle (big) data as well as increased machine learning capabilities, planning functionalities and integrated data modeling possibilities. Also, technology today allows to integrate planning layers and models, thereby enabling end-to-end visibility and synchronized planning. In the near future, the vast majority of activities in supply chain planning will be handled by digital solutions, allowing supply chain planners to fully focus on reviewing the exceptions, responding to events as they happen and collaborating closer with other functions.
How will this improve supply chain planning? First of all, exception-based planning will be end-to-end. If a demand shortage is expected at one location (e.g. DC), an algorithm will assess and provide options and corresponding costs to (1) ship the products from other stock locations, (2) produce missing products additionally vs. (3) lost sales. Second, we are moving from descriptive (notifying the planner about an exception) to prescriptive (notifying the planner about possible solutions to resolve the exception) to decisive, based on self-learning (the usual response of the planner to given options). This will be very helpful when supply chain planning workload increases, for instance when the number of sales orders is increasing.
Near real-time scenario planning
Another pivotal development is that (near) real-time scenario planning is now a reality. The growing computational power and maturing technology offer the opportunity to handle large amounts of data and to run scenarios more frequently, even on the spot. In the case of the COVID-19 pandemic, companies that were already quite advanced in the field of planning technology, were able to react immediately to, for instance, borders that were closing in European countries. Within minutes they knew what the impact of various decisions would be (e.g. reshuffling inventory, and sourcing suppliers). This will become a more widespread capability in the next 10 years. Soon, systems will start to learn from the decisions taken by planners and will eventually make these decisions themselves – based on rules that were defined by supply chain planners in collaboration with various stakeholders inside and outside the company.
Enhanced accuracy of forecasts and planning
In order to build accurate forecasting and planning models, it is necessary to integrate the various internal and external data sources and systems. However, keep in mind that if the data is not reliable, neither is your forecast. This is where new technologies such as AI and machine learning come in. They can help to prepare master and transactional data in order to ensure consistent models based on defined rules.
The amount of data keeps growing, including publicly available data (about e.g. the more accurate and longer-range weather forecasts) that can help to support decisions. For instance, if we look at trade promotion management, one of the current obstacles is the large amount of detailed information that is required. In the near future, the Sales department will only need a few promotional indicators (e.g. which product, which week). The tool will then do a promotion forecast/planning based on historical data about promotions, or even on social media scraping .
Reaching your full potential
In short: there are many digital solutions to automate processes and provide insights. The next 10 years will see the rise of even more opportunities leading to unprecedented high levels of automation. In order to fully benefit from these technologies, we need to start preparing now by taking the required steps towards a shift in mindset and company culture. At the moment, technology is ready, but slow user adoption prevents many organizations from reaching their full potential. For one, people in supply chain planning will need to learn to focus on input (models, parameters and data), as well as on analyses and scenarios, and learn to trust and value these. Also, supply chain planners need to be exception-focused and proactive. In order to have the people with the right skills in place, organizations will need to retrain current staff and hire people with other profiles and backgrounds. This will be the topic for the next article on the Future of Planning: “A planner’s job in 2030 will be nothing like today”.
Examples of automated planning algorithms
There is a growing number of integrated and aligned rule-based planning algorithms across the entire planning cycle to take over the main activities and offer insights to support faster decision making:
- Time series analysis and pattern recognition: machine learning data analysis and insights to improve forecast models.
- Statistical forecasting and demand sensing: providing initial forecast/expectations on a monthly, weekly and even daily basis, reducing manual planning efforts.
- ABC-XYZ segmentation for products/customer combinations: determining for which segments decisions can be made automatically; in an automated way.
- Multi-echelon (Safety) stock determination / optimization and placement based on supply chain variances (reducing human input in determining stock levels).
- Supply optimization for profit- or cost-driven decision making (including indirect costs for e.g. environmental impact and sustainability).
The future of planning series
Now is the time to start preparing for the future of planning – in terms of technology, talent, and organization. The COVID-19 pandemic could in fact be the turning point, if you want your company to respond, recover and thrive in the post-crisis era. In this series of online articles, we will share our in-depth views on these topics.
Article 1: How 5 megatrends will disrupt your supply chain planning.
#1 An increasingly dynamic world. How to speed up decision-making and shape “the new normal”.
#2 Changes in consumer behaviour. The world is individualizing. This trend started in consumer business but is now dripping into other industries as well.
#3 Exponential growth in digital capabilities. The age of exception management and near real-time scenarios has started.
#4 The war on talent. How to attract and retain the future “gems” in supply chain planning?
#5 Purpose-led companies. The focus of companies and employees is shifting towards topics like sustainability. This too, has a huge impact on decision-making, e.g. where to source.
How 5 megatrends will disrupt your supply chain planning?
Go to the article series
For more information about the Future of Planning series or Supply Chain Planning, please do not hesitate to contact Jurg, Leon or André via the contact details below.
The future of planning, part III
Co-create your digital supply network strategy and turn your supply chain into an always-on strategic differentiator