FEATURE
AI in supply chain helps us make better decisions at scale, but human expertise remains key.
Rather than relying on isolated digital applications, the ambition is to connect information across the supply chain so that a change in one part of the operation can inform decisions elsewhere.
World Cup puts supply chain to the test
Few events illustrate the scale of that challenge better than the FIFA World Cup 2026.
Unilever activated 35 brands and 180 limitededition products across more than 120 countries and millions of retail locations as part of its sponsorship.
Behind the marketing campaign sat a substantial supply chain operation involving sourcing, manufacturing, logistics and retail execution.
Planning began by determining which products individual brands would promote and ensuring sufficient raw materials, manufacturing capability and logistics capacity were available.
Once production plans were established, attention shifted towards ensuring products reached retailers at the right moment. Those destinations ranged from major international retailers to hundreds of thousands of smaller independent stores.
At the same time, marketing campaigns ran across social media, television, outdoor advertising and physical stores, supported by more than 50,000 creators globally.
That created the potential for sudden demand spikes, meaning even detailed forecasting needed to be accompanied by operational flexibility.
The response following the tournament final demonstrated the speed required. Within days of Spain winning the competition, special limitededition‘ World Champions’ Rexona deodorants were ready for dispatch.
Such responsiveness depends on connecting marketing activity, consumer demand signals and manufacturing capacity much more closely than would have been possible using traditional supply chain processes.
AI forecasting at massive scale
At the centre of Unilever’ s technology strategy is the ability to make sense of huge quantities of operational data.
Its Forecast Engine Utility combines machine learning and data science to generate a 104- week forecast every week across more than five million product-customer combinations in 40 operating markets.
That scale would be extremely difficult to manage through conventional forecasting methods.
“ AI in supply chain helps us make better decisions at scale, but human expertise remains key,” Cuthbert says.“ Our people provide the judgement, context and experience needed to make the final decisions.”
This human-plus-AI model is central to Unilever’ s approach. AI processes information, identifies patterns and supports scenario planning, but responsibility for interpreting those insights remains with employees.
The technology is also moving deeper into manufacturing.
At Hefei, AI has contributed to an 8 % improvement in overall equipment effectiveness and a 20 % reduction in waste.
Unilever’ s Raeford factory in North Carolina, which played an important role in producing products for the FIFA World Cup 2026, provides another example.
The facility uses digital twins to model manufacturing processes, contributing to improved product quality, a 20 % reduction in waste and a 10 % increase in capacity.
Digital twins effectively provide manufacturers with virtual versions of physical processes. Engineers can examine performance, test potential changes and identify problems without disrupting the real production environment.
Building the connected supply chain
The next stage of Unilever’ s strategy is to move beyond individual AI applications towards an interconnected digital environment.
A five-year partnership with Google Cloud is intended to establish an enterprise-wide digital backbone capable of turning data into actionable insights and supporting agentic workflows across business processes.
Manufacturing is expected to become increasingly predictive and scenario based as part of that transition.
26
INTELLIGENT CIO NORTH AMERICA www. intelligentcio. com