4 most effective uses of generative AI in retail
Discover how the most advanced retail organizations are using AI today and what they’re planning for the future.
A trusted partner can help your retail organization plan and implement an efficient, powerful and secure AI solution.
Generative AI has just crossed into the mainstream for retail IT. In fact, 65% of businesses are either accelerating existing generative AI strategies or creating their first AI strategy.
As more IT teams and business leaders embrace AI and find ways to customize the technology to fit their specific customers, some universally effective and inspiring uses have arisen.
1. Controlling costs is where it all starts
The clearest use for generative AI, and the foundation for most AI strategies, is cost control — which can include everything from productivity and IT equipment to pattern detection and loss prevention.
The savings can be immediate and substantial when applied to administrative and repetitive tasks, freeing up workers for more important initiatives. A Stanford University study found that generative AI could increase productivity in the retail and consumer packaged goods (CPG) industry by up to 2% of annual revenue (or an additional $400 billion to $660 billion).
The prospect of such operational efficiencies is the reason cited by 46% of retail and eCommerce companies for adopting AI technology.
AI is also helping with shrinkage, which has grown into a $100 billion problem according to a survey from the National Retail Federation. Theft accounts for 65% of the losses. The survey also identified a 26% increase in theft attributed to organized crime and reported that 38% of retailers say they’ve seen an increase in return fraud over the past five years.
89% of retailers say they need better analytics capabilities to further improve their loss prevention efforts.
Generative AI algorithms can help by detecting anomalies in patterns from customer preferences to fraudulent transactions — quickly and at scale. This pattern detection can cut down on the number of false positives in fraud as well, which reduces customer service costs and improves customer experience.
These tools can also issue direct responses to customers at checkout to resolve issues much faster than ever before while minimizing the need for manual interventions.
2. Improving customer experience takes personalization
Half of the retail and eCommerce companies surveyed are adopting AI for use in customer service. Generative AI analyzes vast amounts of customer data to generate personalized recommendations, offers and experiences that support the entire purchasing lifecycle.
By analyzing a customer’s preferences in real time, AI can help with smart checkouts, recommendations for additional items and virtual try-on with augmented reality.
It’s already working. 35% of consumers surveyed say chatbot integration increases their likelihood of doing business with a company. This may be why 61% of retailers say they plan to use AI for applications like chatbots.
But the AI is only as good as the data the business feeds it. Generative AI models require continuous human intervention and management to ensure the customer is served helpful and accurate recommendations.
Training complex models requires accelerated computing. Experts like CDW, with partners like NVIDIA, can help manage and deploy these solutions.
NVIDIA is bringing retail into the era of AI with a full-stack architecture powered by NVIDIA AI Enterprise, the end-to-end software for production AI.
Using NVIDIA RAPIDS,™ included in the NVIDIA AI Enterprise platform, retailers can harness the power of data for analytics and achieve faster time to results. When NVIDIA RAPIDS is used with NVIDIA Merlin,™ users can build high-performing recommenders at scale to get better predictions and increased click-through rates.
No matter where you’re starting, explore your NVIDIA AI options today with your CDW account manager or call 800.800.4239.
3. Enhancing data privacy and security protects it all
AI-based tools can help with security by detecting abnormal behavior and fraud and aiding in loss prevention. AI models can do the same for networks.
What keeps AI particularly effective and seamless in a business is protecting the data that makes up the AI. The increased need for quality data has created increased security risks. Cybercriminals value customer data as much as retailers.
According to the National Retail Federation, nearly 59% of retailers have seen an increase in cybercrime over the past five years, and 55% say their loss prevention teams should be more involved in cybersecurity efforts.
With the average cost of a data breach in the retail industry topping $3 million, businesses realize they must first secure their data to implement an effective AI strategy.
Regular monitoring, auditing and testing of AI models is essential not only to protect the data, but to also identify biases, address ethical concerns and ensure the responsible deployment of AI systems within the business.
Government data regulations also apply for most businesses, which may mean the use and location of the data must be restricted. Employ the expertise of a trusted partner like CDW to ensure your data is safe and effective from anywhere.
With NVIDIA Morpheus, part of the NVIDIA AI Enterprise platform offering, developers and customers alike can build optimized AI pipelines for filtering, processing and classifying large volumes of real-time data — bringing a new level of information security to retailers.
4. Innovating at the edge produces the greatest impact
What may hold the most promise for retailers is the implementation of edge computing — including internet of things (IoT) devices at the edge for powerful and responsive insights wherever they’re needed.
Edge computing can lower the operational costs associated with cloud solutions. Businesses can process and filter data at the source, sending only the relevant information to the cloud for further analysis.
Conquering latency and optimizing connectivity is the key to making retail IoT devices effective. For example, retailers who use radio frequency identification (RFID) technology have seen a more than 25% improvement in their inventory accuracy and an over 60% improvement in their profit margins. Making sure that data is usable in real time with edge computing is crucial.
The compute power of the edge is being put to good use with the 30% of retailers who say they’re implementing or planning to implement point-of-sale video analytics. In combination with IoT sensors, businesses and customers can get immediate feedback when checking out, based on cart contents and pattern detection.
Behind the scenes, 53% of retailers say they plan to use AI for product and attribution data, as well as for forecasting, purchasing, pricing and inventory management. NVIDIA is helping retailers drive these decisions in real time, taking their AI compute to where the data is — at the edge.
Expertise will be a required part of this strategy to keep the network optimized and the infrastructure scalable so future innovations can be built on the edge.