NVIDIA survey finds 90% of retailers are adopting or piloting AI

Retail has officially moved past the stage of slapping a novelty chatbot onto an online storefront just to impress shareholders.

In an industry known for razor-thin margins, technology either pays for itself on the ledger or gets cut immediately. NVIDIA’s second annual State of AI in Retail and CPG survey makes it clear that artificial intelligence is now firmly embedded in everyday store operations, inventory planning, and supply chains.

Based on responses from hundreds of industry professionals, the survey found that 89% of retail and CPG companies are actively running AI in their operations or running pilot programs, up from 82% in 2023. Even more telling: 87% reported that AI drove positive annual revenue growth, and 94% said it helped reduce operating costs. With returns like that, 97% plan to boost their overall AI spending in the next fiscal year.

Marketing was the gateway, but ops is the real prize

Generative AI adoption has moved remarkably fast. Over 80% of companies surveyed are now deploying or testing generative models, and 93% plan to expand their genAI budgets next year despite persistent concerns around data privacy, security, and implementation costs.

Right now, content generation for marketing leads the genAI pack at 60%, followed by predictive analytics at 44%, personalized marketing at 42%, customer segmentation at 41%, and digital shopping assistants at 40%. More than half of the retailers surveyed are now juggling AI across more than six distinct business workflows.

Physical stores are seeing some of the most practical deployments. In brick-and-mortar locations, the top use cases are inventory management, store analytics, and adaptive advertising. Meanwhile, supply chains—where 59% of respondents noted mounting logistical pressures over the past year—are rapidly turning to AI. Around 58% credit AI with boosting operational throughput, while 45% use it to slash logistics costs, prompting 82% of brands to plan increased supply chain AI investments.

Yet friction remains. The single biggest hurdle retailers cited in 2024 was a shortage of easy-to-understand, explainable AI tools. Black-box models simply make floor managers and merchandisers nervous when million-dollar inventory calls are on the line.

Why it matters

We think this report marks a decisive pivot from generative AI as an experiment to generative AI as baseline enterprise infrastructure.

For retail operators and consumer brands, the playbook is clear: high-margin, low-risk use cases like marketing asset creation and automated merchandising recommendations deliver almost immediate payback. If you are not testing them today, your competitors are already using them to shave points off their operating margins.

However, the widespread complaint about explainability reveals a crucial bottleneck. Retailers are willing to hand over creative drafting and predictive trend spotting to AI, but they hesitate on mission-critical inventory decisions without clear rationale. The AI vendors that win the next wave won't just offer raw model power—they will deliver explainable agentic workflows that retail staff can actually understand and trust.