Business leaders aim to build predictability into every aspect of operations, especially the supply chain. Traditionally, this means investing heavily in data and analytics, hoping the past will inform the future. 

It often doesn’t. As a result, we spend countless hours in meetings reviewing complex spreadsheets and lengthy presentations that convey only numbers. The “why” is lost in analog, subjective and disparate process. 

Supply chain decision-making is often misaligned. You might, for example, see a sales team budget for 30 widgets. Yet finance is told that they only expect to sell 20, and supply chain that they need 40, to ensure sufficient stock.  Each department has its own goals and, sometimes, its own political agenda.  The result is an inefficient process where teams spend more time defending their numbers than aligning on the best outcome.

Artificial intelligence presents an opportunity to redefine planning and bridge the gap in decision-making. It enables businesses to connect data to the real world and create insights that are visible to everyone in natural language. Instead of a static spreadsheet, Through creation of “digital insight,” AI can surface the drivers behind changes to the plan, complete with explanations.

The most effective approach to implementing AI-enabled planning is to begin with one or two key processes. For leaders starting this journey, following are five best practices:

Establish a clear leader, with end-to-end responsibility and accountability. In siloed organizations, process management often prioritizes departmental key performance indicators over customer needs or organizational goals. This new owner must have the authority to speak for their department, but also the mandate to advocate for the interests of the entire organization.

Identify process users and their needs. Applying AI to demand forecasting and supply chain predictability requires clear processes and outcomes. The demand plan must be accurate and include a clear understanding of deviations. The supply plan emphasizes stock levels and delivery, while finance requires a higher return on capital employed. The new process must work with a decision hierarchy that is agreed upon by all parties.

Evaluate your current execution. Before redesigning a process, you need to understand how information currently moves through your organization. Start by creating a high-level flowchart. These were once built with sticky notes and whiteboard drawings, and digitizing the steps is simple with the right systems. The real challenges often lie in the undefined “white spaces” between departments. These gaps are where communication is unclear and no single team owns the responsibility, and are frequent sources of delays, errors and inefficiencies.

Form a cross-functional redesign team. This should comprise representatives from all relevant departments. The team must be able to identify the new skills, technologies and changes to performance metrics and organizational structure that are needed to support the new process.

Put people at the center. There’s the common fear that AI will replace jobs, but this new approach does the opposite. It moves humans to the center of new, machine-assisted processes. Supply chain predictability is still a human-centered priority. People set goals, and AI helps to achieve them with more skill and precision than ever before. 

If making supply chains entirely predictable were possible, we would have achieved it already. AI can't promise perfect predictability, but it can create clarity. And it can provide a way to manage volatility. However, improving this important business process isn’t as simple as just plugging in an AI application. We need to rethink how we plan, align and communicate.

We must stop defending our numbers and start collaborating on our insights. The organizations that thrive will be those that learn to plan, align and communicate in this new, data-driven language.

Piet Buyck is senior vice president, innovation strategies at Logility, and author of  AI Compass for SC Leaders.