Underlying models: Best-fit and blended models powered by AI (analytical intelligence) can improve (and have improved at several companies already) accuracy by 15-30%
Learning from extra input and data sources: AI processes more data than traditional methods, improving accuracy by 10-20% (contextual and analytical intelligence)
Generating insights and prioritizing tasks: AI helps planners save up to 70% of their time by automating key tasks (user intelligence)
AI is transforming demand planning by increasing forecast accuracy, reducing the workload for planners, and freeing them up to focus on collaboration and strategy.
AI-powered models allow businesses to analyze more data, learn from historical and external inputs, and generate actionable insights much more effectively. Here’s how AI influences the different aspects of demand planning:
AI tools can automatically select the best-fit model for different product categories based on historical performance. Traditionally, planners had to test multiple statistical models manually. Now, AI can run all the models simultaneously and determine the optimal one for each situation.
Best-fit models: AI analyzes various models such as ARIMA, exponential smoothing, and various machine learning methods, and picks the one that delivers the highest accuracy for a given product and time period.
Blended models: AI also enables the use of blended models, combining multiple approaches to capture different elements of demand. For instance, one model might be better for baseline forecasts, while another handles promotions or seasonality more effectively. This hybrid approach enhances forecast accuracy across a range of products and categories.
One of AI’s strengths is its ability to process amounts of unstructured data like CRM data or even news reports to generate alerts. AI continuously learns from past demand fluctuations and integrates that knowledge into future forecasts, making them more accurate over time.
Large Language Models (LLMs) are starting to play a major role in demand planning workflows by generating insights, automating tasks, and helping planners prioritize their work.
Workflow intelligence: LLMs automate routine tasks like data gathering, analysis, and reporting, saving planners time and allowing them to focus on higher-level decisions. They simplify complex data by presenting key takeaways that matter most to the planner.
User intelligence: AI tools provide intelligent dashboards and customizable reports, giving planners and stakeholders real-time access to forecasts and insights. This not only improves alignment across departments but also makes it easier to adapt strategies based on the latest demand planning assumptions.
Interesting demo video by Logility of their demand planning environment with a chatbot assistant that’s using LLM to generate reports and answers to questions.