Assortment Planning & SKU Optimization

Effective pricing and quoting begin with the right products on the shelf. The global assortment and space organization market is projected to grow from $2.06 billion in 2024 to $4.92 billion by 2033, according to MarketsandMarkets. The expansion underscores the growing complexity retailers face as they juggle consumer demand shifts, localized preferences, and SKU proliferation.

Only a fraction of shoppers remains loyal when faced with an empty shelf. A 2024 report from supermarket chain Kroger’s data analytics firm 84.51° found 23% of shoppers will search elsewhere if they find a grocery item is out of stock and 4% will move their entire shopping cart to another retailer. For retailers managing tens of thousands of SKUs, balancing national brands, local preferences, and seasonal trends has become an intricate exercise in precision.

AI is transforming assortment planning from a reactive, intuition-driven process into a predictive, data-led discipline. Machine-learning models now analyze sales patterns, anticipate demand shifts, and automate inventory allocation. These tools integrate predictive analytics, computer vision for product attribute recognition, and natural language processing to assess customer feedback. By incorporating both internal sales data and external signals—such as weather forecasts and social media sentiment—AI enables planners to make localized decisions in real time.

Advanced clustering algorithms like principal component analysis help retailers group stores by behavioral and demographic characteristics rather than size alone. Transferable demand models forecast how shoppers might respond to SKU additions or removals, identifying substitution patterns that manual methods miss.

Yet technology is only part of the equation. Successful AI adoption depends on robust data governance, strong integration with enterprise resource planning (ERP) and supply chain management systems, and disciplined change management. Retailers must replace spreadsheet-based planning with connected platforms that unify sales, margin, and customer metrics.

Human readiness remains essential. According to a recent McKinsey survey, 86% of grocery executives view operational efficiency as AI’s greatest benefit—but realizing those gains requires retraining and process redesign. Category managers must learn to interpret algorithmic recommendations and understand when to override them to align with brand and vendor agreements.

Results from early adopters show significant value. A European retailer identified 150 high-value SKUs that deserved more visibility and 200 low-value ones that deserved less, cutting €2 million ($2.17 million) in operating expenses within 90 days, according to Throughput Inc., the provider of the AI technology. The retailer further projected up to €10 million ($10.86 million) in profit improvements per facility through optimized SKU allocation. A McKinsey case study found assortment optimization can reduce SKUs by 36% while still growing sales and margins by up to 2%. One consumer goods company eliminated 40% of low-performing products in a single year, improving cost structure and focus.

In B2B distribution, the fiscal impact is equally strong. Studies show that AI-driven demand forecasting has reduced inventory by 20% to 30% for some industrial distributors while improving fill rates by up to 8%. A leading building- products distributor used an AI-powered supply chain control tower to enhance availability, while an industrial supplier applied generative AI to uncover $2 billion in sales opportunities. A global petrochemical firm gained about $100 million in earnings through machine-learning-enabled dynamic pricing. 119 2.2 Sell (Conversion & Revenue Growth) Grocers, which typically sell many thousands of products in each store, expect to quadruple AI spending by 2025, prioritizing demand forecasting and localized planning. Retailers using advanced analytics and market basket analysis report gross margin gains of 2% to 3%. Most start with pilots in limited categories, then expand after refining their models and governance processes.

The technology ecosystem has matured rapidly. Looking ahead, generative AI will automate even more of the process—drafting product descriptions, analyzing reviews, and generating “what-if” scenarios for planners. Combined with computer vision shelf monitoring and Internet of Things (IoT) inventory sensors, the future points toward fully autonomous assortment management.