The Supply Chain Planner of the Future

Advances in supply chain 4.0 technologies have triggered visions of managers across the globe driving automation to the next level. Take Richard Liu, the founder of the Chinese retail giant JD.com, who described for Forbes his vision for full automation within his company with “no human beings anymore, 100% operated by AI and robots.” This is what we call a true no-touch supply chain.

Some of this is already happening. Technologies that enable automation of the physical flow of goods are currently being widely piloted, including lights-out factories and warehouses, piece picking robots, automatic-guided vehicles, early iterations of self-driving trucks and drone delivery.

Similarly, ordering processes are also advancing, quickly moving away from fax and phone to the Internet, EDI, personal assistants/chatbots and IoT—automation degrees of more than 90% are possible even in complex B2B environments.

The benefits of automating physical flows are obvious: 24/7 operations, reductions in lead times and errors, consistent processes and real-time optimization on the fly.

Many interesting technologies are also coming together to automate information flows and decision-making, including Cloud platforms, robotic processes automation (RPA), negotiation bots and artificial intelligence.

Accordingly, the outlook for many blue and white collar jobs has dimmed significantly, at least according to some analysts who predict a future where robots and robotic process automation will replace a significant percentage of today’s routine, repetitive jobs.

Supply chain planners may wonder how all of this will affect their jobs. Will technology make them obsolete too? In our view the impact of automation is potentially high for operational processes, especially repetitive processes, but the impact is lower for processes with strategic importance that benefit from increased data availability and advanced analytics to support manual decision taking. We put supply chain planners in that lower impact camp. Figure 1 provides an overview of the automation potential landscape.

Figure 1. Automation potential landscape

Today’s requirements for planners

Planners can be found in a wide range of supply chain functions, including demand planning, inventory planning, supply planning, production planning, distribution requirements planning, order management and more. They typically account for one third of administrative roles in supply chain management—more than any other function.

In a recent study in Supply Chain Management: An International Journal, 243 experienced supply chain managers participated in a choice tournament to identify the hiring criteria for planners.

In the experimental setting, multiple alternative candidates were repeatedly presented to managers who then decided on their preferred supply chain candidates. The results indicated the different priorities on the competence profiles of planners that exist.

While IT competencies and industry experience played a similar role across all participants, the importance of supply chain expertise and interpersonal skills differed widely.

Only 38% of managers focused very much on extensive supply chain management knowledge and were characterized as “expert chasers;” in contrast, 62% of managers preferred candidates with very good interpersonal skills, such as being a team player and a good communicator, and with high problem-solving abilities, such as critical reasoning and conceptualization.

While the study gave no clear indication which planners fit better in which situations, it seems obvious that the results could be linked to the maturity of planning that differs widely across companies.

Automation of planning is still in its infancy in less advanced companies that use tools from the 1990s and deal with master data problems from the 2000s. Here, planning is manual and time-consuming, and changes are cumbersome.

A lot of interaction with other functions in the supply chain is required to align requirements and meet tight deadlines. More advanced companies have fit-for-purpose systems and are piloting new, state-of-the-art planning solutions.

Many planners work for companies that fall in a range from those that are gradually adopting new technologies to those with moderate maturity levels. Still, there are examples of companies that are already far ahead and are experiencing now how the role of planners will be affected by new technologies and advancing process maturity.

So, what will change for planners? We anticipate three significant shifts: First, in general, planning tasks will shift in more mature companies; second, multiple dedicated roles will be created around planning; and finally, new organizational setups will be adopted.

Planning tasks will shift

Planning has always been challenging. Planners are constantly filling the gaps in broken business processes hindered by legacy IT systems, limited system integration and poor master data management. Further, planners have often been overwhelmed by the number of decisions that they need to make with increasing numbers of products, higher product complexity, frequent changes and lack of sufficient buffers.

Often, they have to go through a long list of products one-by-one to decide on quantities and timing. While many supply chain planning systems automatically generate demand forecasts and production/purchase order recommendations, these numbers are frequently not trusted and manually checked or even overruled.

New technologies and solutions around Robotic Process Automation (RPA) and predictive analytics promise to avoid the need for much of the manual and routine work. RPA has been on the rise during the last few years as more and more companies experiment with the potential of automating tedious, manual processes.

In our experience, repetitive tasks currently done by planners can be automated in approximately two months from the initial assessment to roll-out, whereas previous automations within an ERP system typically took 9 months to 18 months. The cost of this type of automation is much lower, in the range of $10,000 to $30,000, depending on the complexity of the solution, so that smaller processes can be efficiently automated.

Predictive analytics in demand planning is another example of how technology reduces manual planning efforts. In well-set-up demand planning solutions with optimized time-series-based statistics, most manual overrides to the statistical forecasts are driven by the need to incorporate additional information not seen in the past demand history.

Examples of this might include information about upcoming promotions or new points-of-sale, but also customer-specific information such as new contracts or signals of anticipated demand.

These can be efficiently covered through predictive analytics, where multiple input sources can be combined and assessed to form a basis for the anticipated future demand. In one example, a beverage producer used predictive analytics to cut the efforts for manual promotion planning by over 80% while increasing forecasting accuracy.

Only the type of promotion (2-for-1, 3-for-2, price decrease) was manually provided by the commercial team, and machine learning approaches were then used to assess the expected impact on sales. At the same time, the quality of the forecast was significantly improved (forecast error on weekly basis reduced by ~50%), which justified the automation approach.

Figure 2. Implementation priorities for automating order decisions

With the introduction of these technologies comes a task shift for planners. While not all planners will need to become data scientists, they will benefit from automation using supply chain data provided by data scientists. These experts will develop and maintain solutions that automatically check the data, identify issues and ensure a high-quality planning basis. Accordingly, planners will spend less time on manual data inspection and constant fire fighting to plan and schedule the products manually.

Similarly, Artificial Intelligence (AI) forecasting approaches are able to more accurately forecast standard demand and allow planners to focus on fewer items that really do require attention. Smart algorithms detect items where human input is required for better decision making, such as when historic demand is not a good indicator of future sales, and when information on market developments and pending offers need to be integrated manually. The advanced planning systems point the planners toward those exceptions and help to focus their efforts on where they are really adding value. Some of the key task shifts are outlined in the following.

• Demand planning. So far, demand planners have been focusing on managing the end-to-end demand planning process, covering selection of the best-fitting statistics, managing the manual validation process, analyzing effectiveness of forecasts and quality of inputs and preparing and facilitating demand validation meetings. Predictive analytics can change the way that planners operate. In the future planners will focus much more on providing the right input data sources and handling exceptions such as new product introductions or end-of-life planning. The actual analytics will very often be outsourced and be run almost as a “black box.” Based on their performance, analytics solutions gain the trust and commitment of the planner over time.
• Production planning. The current tasks of production planners are often focused around matching demand and supply and manual re-planning. State-of-the-art planning solutions are already able to optimize production for margins and to re-plan given certain triggers in execution. The main benefit of the supply planner will evolve to maintain and improve the optimization logic of the planning systems (think adapting the objective function or the constraints to better resemble the business objectives and the operational reality) while also ensuring that the data basis is up-to-date in accordance with the stakeholders from operations and commercial.
• Distribution requirements planning. Today, planners typically have to review hundreds of articles across multiple systems, such as the advanced planning system and ERP, and potentially even across multiple instances of these systems to forecast demand, replenishment and consequent stock development. Planners review the system-generated purchasing requisitions and manually decide if they should be converted to firm purchasing orders (again see sidebar: Automating order decisions). While the system logic for creating purchase requisitions is well proven and standard based on MRP, the main reason for manual deviations from system-generated proposals is the data quality or missing information in the system. Going forward, planners will focus much more on maintenance of the master data and exception handling in case of allocation situations. The previous enables a no-touch replenishment planning based on predefined rules.

New roles will be created

Many planning roles are now structured along the planning steps of demand planning, supply planning, production planning or scheduling. In the future, planning roles will evolve from a domain-focus to attention to tasks that require specific capabilities (and can be used across all planning domains). In particular, planners will specialize in data management, algorithmic optimization, exception management and partnering with the business, creating roles around the following tasks.

• Data management. The right data is the basis for any planning automation, which entails both data availability and data quality. This will benefit from dedicated supply chain analysts and data engineers who set up procedures and RPA protocols to automatically check for data gaps and consistency by comparing master data with actuals to identify and correct for deviations. Another crucial task is to design and align data exchange with supply chain partners. For example, Amazon uses innovative EDI systems and APIs for connecting to vendor systems and educates the vendors on how to best use these interfaces to ensure seamless connectivity and high data quality. To enable drawing the right conclusions from the data, data visualization engineers fill another core task within data management. They develop new perspectives on visualizing the data and develop customized user experiences and interfaces that enable planners to understand those complex data sets.
• Algorithmic optimization. As supply chain planning will be widely automated, the key differentiator will be the optimization quality of the algorithms. To build out this advantage, there is a need to continuously optimize the analytics using approaches around advanced modelling, machine learning and stochastic optimization. Planners will need to combine cutting-edge research with pragmatic incorporation of learnings into the planning algorithms—similar to what leading e-commerce players already do today. For example, Zalando combines algorithms developed and described in academia and adapts these in a pragmatic trial-and-error approach for use in their warehouses. The quality and performance of the algorithms are then tested in limited pilots within the live system to prove their effectiveness.
• Exception management. While algorithms and software robots will take over many routine activities, exceptions will still occur and conflicts will need to be resolved. To do so, exception managers need to evaluate scenarios and options (supported by the system), and make trade-off decisions together with internal and external stakeholders. This role requires good communication skills to align the needs of all parties involved. The role of the exception manager is a good example of the “automation paradox” that describes how the automation of simple tasks creates new issues as employees now constantly deal with difficult cases.
• SCM business partner management. The shift toward more capabilities-centered roles requires a translator role within supply chain management to act as a counterpart to the business and operations—to translate their requirements and reality into the planning process.  This “SCM business partner” will play a pivotal role in shaping the design of the supply chain planning processes and systems. Also, the SCM business partner will be driving and optimizing the S&OP/IBP process, to facilitate between the functions and find an overarching business optimum.

New organizational setup will be adopted

The next level of automation and the new roles in supply chain planning do not primarily aim at reducing the cost of planning. Rather, the new setup will cut lead times and increase customer service level, minimize errors, optimize margins and provide a real competitive advantage.

For this to work, however, it will require a new organizational setup for planning. As a large share of routine planning activities are automated, the physical proximity to operations (production planners being located in the plants) will be less important while a knowledge exchange and best practice sharing among planners becomes even more crucial. Thus, planners will require a co-location to enable the continuous interaction and improvement.

Figure 3. Evolution of planning setup

We might therefore see a resurrection of shared service centers, or shared planning hubs, for supply chain planning—this time, however, to move closer to the right talent and to enable close collaboration, rather than to save personnel cost. While the number of planning FTEs will probably go down, the total cost will most likely not change dramatically, as the change of roles comes with increasing capability requirements and higher salaries.

The need for highly qualified talent is continuously increasing and already today, there is a tendency to outsource such jobs into shared service centers with access to the right talents rather than into low-cost countries. The prime decision criteria will be how to best access the top talent in data engineering and algorithmic optimization.

The next planner

Future planning will not make planners obsolete, but the “planner 4.0” will not be comparable to many of today's planners. Planners need to become much more analytical and IT-savvy, requiring significantly enhanced capabilities—and will be a core differentiator for supply chain performance.

Already, companies can prepare for future requirements by establishing the data engineering and optimization roles for creating batch jobs doing “automatic” master data checks and creating suggestions for adaptation. While doing so, it is essential that companies codify the extensive experience and knowledge of their planners and ensure that future systems can use it rather than having to rediscover core aspects of the operational reality. What’s more, skills in supply chain planning and advanced analytics need to be urgently built to not lose ground in the context of a fast-evolving supply chain planning function.

Knut Alicke is a partner in McKinsey’s Stuttgart office and leader of the Supply Chain Management Practice in Europe.

Kai Hoberg is a professor of supply chain and operations strategy at the Kühne Logistics University in Hamburg.

Juergen Rachor is a senior expert in McKinsey’s Supply Chain Management Practice.

Related Resource

Are You Prepared For The Supply Chain Talent Crisis?
The industry can build an adequate supply of talent in a number of ways, but companies must be more proactive in their approach to recruiting, developing, and retaining the supply chain professionals they need to stay competitive. Download Now!

 

 

Visit the Supply Chain 24/7 Education Center!