Upsell, Cross-Sell & Substitutions

Once a customer fi nds what they are looking for, the opportunity for upselling and cross-selling begins. Many retailers fail to capitalize on these potential gains because their recommendation systems are inadequate. Managing thousands of SKUs while predicting individual preferences creates an operational burden that traditional tools cannot meet.

McKinsey has found that cross-selling can lift sales by 20% and profi ts by 30%. In business-to-business (B2B) environments, where buyers frequently reorder or purchase for projects, timing and availability are critical. If an item is out of stock, intelligent systems must instantly suggest complementary or substitute products. Without these capabilities, distributors risk customer defection and smaller basket sizes.

Stockouts are a chronic challenge for B2B distributors managing large catalogs. With AI-enabled sales tools, a customer service representative could identify in-stock alternatives and save the sale. Without them, a missed item can easily lead to lost revenue and a frustrated buyer.

Modern recommendation engines rely on machine learning models that combine collaborative fi ltering, deep learning, and real-time personalization. They use three main methods:

• Collaborative filtering, which analyzes user behavior to recommend products chosen by similar customers.
• Content-based filtering, which uses product attributes to find comparable items.
• Hybrid models, which merge both approaches for greater accuracy.

Integrating recommendation technology requires careful planning. Companies can adopt standalone engines or customer relationship management (CRM) systems with built-in AI for upselling and cross-selling, but success depends on clean, standardized product data.

AI systems also depend on historical data, which can disadvantage new entrants. Privacy regulations may further limit data use, and algorithms often struggle to capture complex B2B purchasing dynamics, where multiple stakeholders influence decisions.

Amazon’s AI recommendation engine remains the gold standard, fueling about 35% of all purchases, according to McKinsey. The same logic applies in distribution. For example, a lawn and garden parts distributor generated 21% more revenue per customer by using an AI tool that predicted when customers would reorder.

Telecommunications provider Vodafone boosted its cross-selling success rate through personalization. Across the software-as-a-service (SaaS) sector, 44% of companies report earning 10% additional revenue from upselling and cross-sells. A HubSpot survey found 88% of sales professionals upsell, 79% cross-sell and 72% attribute between 1% and 30% of company revenue to those tactics.

Beyond immediate revenue gains, recommendation engines strengthen loyalty. An Accenture survey found that 91% of consumers are more likely to shop with brands that recognize and provide relevant offers. Businesses using these systems often see fewer service inquiries, faster checkouts, and improved inventory turnover.

Deloitte estimates that applying generative AI to sales enablement, quote generation, and post-sales support can add 75 to 100 basis points of earnings before interest and taxes for the average wholesale distributor.

As recommendation systems merge with GenAI, new capabilities are emerging—such as natural-language product comparisons, intelligent substitutes, and conversational search.