Canadian retailers enhancing customer experience through generative AI

The most common uses for generative AI among retail respondents include:

• Detecting fraud by raising red flags for suspicious transactions (69%)
• Predicting product demand, optimizing inventory levels (68%)
• Offering personalized product recommendations in customer-tailored conversation styles (67%)
• Powering product search engines by making it easier to understand customer search inquiries (67%)
• Inbound and outbound Scheduling/Truckload Optimization (67%)

"It is clear Canadian retailers see generative AI as critical to their futures," says Kostya Polyakov, partner and national industry leader of KPMG in Canada's consumer and retail practice. "The challenge is identifying use cases that add value to organizations since there are a myriad of ways retailers can use the technology to become more efficient, productive and profitable."

READ:  Ask the Expert: How can AI surveillance modernize inventory management for c-stores?

"Generative AI is a natural fit for retail: it can personalize the customer experience, allow more precise forecasting and improve the supply chain. More than 80% of Canadian retailers we surveyed will be utilizing the tech this year. For the 20% not there yet, they face a huge competitive disadvantage that will only grow as those using it find more and more ways to leverage its powers. Generative AI is not something for the future—it is now table stakes for Canadian retailers," adds Polyakov.

While uses of large language models will continue to grow, it will be important for Canadian retailers to adopt a responsible use of generative AI framework inside their businesses to ensure this technology is used responsibly.

Key survey highlights:

• 38 % of respondents have a generative AI solution already in place
• 39% are planning to implement their first generative AI solution in the next six months
• 17% are planning to implement generative AI in the next 12 months
• 90% agree generative AI is helping to or will help grow their company's revenue and/or market share growth
• 39% expect generative AI to boost revenue by six to 10%; 26% see a 10 to 15% revenue boost; and 23% see a 3 to 5% gain
• 42% expect generative AI to improve sales return on investment (ROI) of between 10 to 20 %; 41 % expect a boost of between 6 to 10%.

Nearly nine in 10 (86%) respondents said generative AI can help to better inform their marketing campaigns and personalize shopping experiences for customers, and 88% agreed the technology can create stunning visuals for product launches and reduce photography costs. However, nearly eight in 10 (78%) expressed concern about how consumers would respond to AI-generated imagery, something Mr. Polyakov says retailers need to be mindful of in addition to the continually developing rules around intellectual property in light of generative AI.

"Consumers are increasingly using generative AI tools themselves, and many are savvy enough to be able to spot AI-generated material. Retailers need to think carefully about how they're using the technology, because it could create reputational, legal and financial risks if not used properly and responsibly. Having proper guardrails and controls around the technology is a must," says Peter Hughes, national customer experience practice leader, KPMG in Canada.

Generative AI in the supply chain

While respondents reported already using generative AI numerous ways in their organizations, less than half (46 %) have applied the technology within their supply chains, with 34 % of respondents planning to implement it in the future.

Of respondents using or planning to use generative AI in their supply chain, four in 10 (43%) said their primary reason is to unlock prescriptive analytic capabilities for customer or sales order fulfillment, such as tapping in-house and external data to identify SKUs and make recommendations to category managers to adjust pricing, promotions, assortment, and delivery and provide mitigation options. Other major drivers include: the ability to analyze information across disparate systems (35%); generating accurate sales predictions based on historical data, trends, seasonality (34%); and inventory optimization (34%).

"Generative AI has the potential to revolutionize supply chain management, logistics and procurement, but only if it's underpinned by reliable, quality data—that's where many organizations face challenges. Their data is not managed and organized in an optimal way," says Polyakov.

Indeed, two-thirds of respondents said one of the main challenges to implementing AI is having non-validated, inaccurate data inputs, which, if used to train the large language models that underpin generative AI platforms, could potentially lead to "hallucinations" or inaccurate or misleading outputs. Seven in 10 (71%) of respondents said their inability to access or leverage data is also a challenge in implementing generative AI.