With the volatile landscape shaping modern procurement, the margin for error in procurement has narrowed significantly. Traditional demand forecasting, long dependent on historical averages and linear human-plus-spreadsheet calculations, is increasingly inadequate for navigating global disruptions. The integration of artificial intelligence (AI) into demand forecasting represents a shift from reactive purchasing to predictive orchestration.
Moving beyond linear models
Historically, procurement teams relied on time-series analysis, projecting future needs based on past performance. While functional in stable markets, these models struggle with black swan events or rapid shifts in consumer behaviour. AI-driven forecasting, specifically through machine learning (ML), allows organisations to move beyond internal historical data.
Modern algorithms now synthesise vast datasets, incorporating external variables such as geopolitical shifts, weather patterns, shipping delays and even social media sentiment. By identifying non-linear correlations that escape human analysis, AI provides a more granular view of future requirements.
The mechanics of predictive procurement
The strength of AI in this field lies in its ability to handle multivariate analysis at scale. Where a category manager might consider three or four variables, an ML model can assess hundreds simultaneously. This leads to several operational advantages:
- Inventory optimisation: By narrowing the gap between forecast and reality, firms can reduce safety stock levels without risking stockouts. This frees up working capital and reduces warehousing costs.
- Waste reduction: For industries dealing with perishables or short-lifecycle products, precision is a sustainability imperative. Accurate forecasting directly correlates with a reduction in discarded inventory.
- Supplier collaboration: High-fidelity forecasts allow procurement leads to provide suppliers with more reliable long-term commitments. This transparency often leads to better pricing structures and prioritised service during periods of scarcity.
The data quality hurdle
Despite the technical prowess of these systems, the output is only as reliable as the underlying data. "Garbage in, garbage out" remains the primary challenge for CPOs. Implementing AI requires a rigorous approach to data hygiene, ensuring that information from ERP systems, warehouses and external partners is standardised and cleansed.
The black box nature of some advanced neural networks can create a transparency gap. Stakeholders are often hesitant to trust an automated forecast that contradicts their intuition. Consequently, the industry is seeing a rise in explainable AI (XAI), which provides the rationale behind a specific prediction, allowing procurement professionals to validate the logic before committing to high-value orders.
The human-in-the-loop requirement
While the computational power of AI is undeniable, the most successful implementations are those that maintain a "human-in-the-loop" strategy. AI excels at processing data and identifying patterns, but it lacks the contextual nuance to understand sudden policy changes or the intricacies of personal supplier relationships.
The role of the procurement professional is evolving from a data gatherer to a strategic validator. The AI provides the data-driven foundation, but the human provides the strategic oversight, ensuring that the forecast aligns with broader organisational goals and ethical considerations.
AI in demand forecasting is no longer a peripheral experiment; it is becoming a core component of resilient procurement operations. By embracing algorithmic precision, organisations can mitigate risk, enhance efficiency and transform the procurement function from a cost centre into a strategic engine for growth. The challenge now lies not in the availability of the technology, but in the readiness of the data and the adaptability of the workforce.
Hormel Foods
Hormel Foods, the global producer behind brands such as SPAM and Planters, has overhauled its supply chain operations by adopting o9 Solutions’ AI-powered platform. Managing a network of thousands of items, ranging from shelf-stable goods to perishables, Hormel faced significant hurdles including seasonal demand spikes and complex multi-tier distribution.
Working with Accenture to deploy the o9 Digital Brain across more than 70 sites, the company has transitioned from a supply-led to a demand-driven value chain. Chakri Gottemukkala, Co-Founder and CEO of o9, notes that the transition has "replaced disconnected tools with a unified enterprise model that links demand signals to supply, inventory and deployment decisions".
The integration facilitates "touchless" forecasting, which improves accuracy for seasonal items and reduces the requirement for manual overrides. For Will Bonifant, Chief Supply Chain Officer at Hormel Foods, the shift is transformative: “By connecting demand, supply and inventory decisions in one streamlined platform, we are shifting from reactive problem-solving to more proactive, data-driven planning.”
Tesco
Tesco has formalised a three-year strategic partnership with French startup Mistral AI, becoming the first major UK retailer to align with Europe’s leading frontier AI firm. The agreement establishes a joint AI lab where Tesco’s technology experts will collaborate with Mistral’s engineers to develop generative AI solutions tailored for complex retail operations.
The partnership aims to enhance baseline forecasting and logistics capabilities through sophisticated data analysis and automated document drafting. Ruben Lara Hernandez, Tesco’s Data, Analytics & AI Director, emphasises that “this agreement builds on our extensive track record in developing new technology and AI solutions, in ways that ultimately benefit our customers, colleagues and suppliers”.
For procurement leads, the integration offers the potential for more controllable and customisable forecasting tools. Marjorie Janiewicz, Chief Revenue Officer at Mistral AI, explains: “Our Applied AI team will collaborate with Tesco’s experts to build customisable, controllable and frontier AI products to improve internal workflows and Tesco's customers’ experience.”
Costco
Costco Wholesale is integrating AI across its global supply chain to maintain its low-cost membership model amid macroeconomic uncertainty. Central to its strategy is the deployment of predictive demand forecasting and real-time inventory tracking, aimed at minimising waste while ensuring high product availability across its warehouses.
Unlike retailers that prioritise rapid digital-only growth, Costco’s implementation is a phased, disciplined integration into its established brick-and-mortar operations. By utilising machine learning algorithms for demand planning, the firm has improved its purchasing agility, allowing for more precise restocking decisions. This is supported on the ground by robotics and automated guided vehicles (AGVs) within its distribution centres, which streamline the movement of goods and reduce labour overheads.
The impact of these initiatives is reflected in the company's 2025 performance data, which showed a 6.8% rise in total net sales, supported by an 11.6% growth in e-commerce. Beyond sales, the AI-driven approach is a pillar of Costco’s sustainability mandate, specifically in reducing perishable food waste through sharper inventory turnover. As the firm prepares to open nearly 30 new warehouses this year, these AI tools will serve as the foundational architecture for managing expanded global inventory complexities.


