Key takeaways:Â
- AI will not replace specialty retail associates but will shift their role from cashier to clienteling expert, stylist and brand ambassador by 2030.
- Pandora implemented AI-powered product search in September and saw product lookups become 66% faster with item-to-cart speed doubling.
- Retailers using unified commerce platforms see stronger AI results because clean, consolidated customer and inventory data enables a complete view of every transaction.
AI may be reshaping retail, but it is not going to replace the store associate, at least not in specialty retail. That is the central argument Sam Kliger, founder and CEO of KWI, made during this week’s Retail TouchPoints Trendcaster webinar. With four decades of experience building retail technology, Kliger offered a direct counterpoint to the automation headlines dominating the industry conversation.
“I can tell you unequivocally, I don’t believe it’s replacing associates in specialty retail or will be replacing associates in specialty retail,” said Kliger. “The real question is, will retailers actually transform their businesses and use AI to empower people or just cut them?”
Watch a replay of the webinar here.
History Suggests Redefinition, Not Elimination
Kliger grounded his argument in retail history, pointing to two earlier technology shifts that were predicted to eliminate jobs but ultimately redefined them instead.
When ecommerce emerged in the late 1990s, the prevailing narrative was that online shopping would eliminate physical stores. It did not. Stores that lacked strong associates, relevant product and sound operations struggled. But stores that invested in people and experience continued to perform. According to Kliger, approximately 80% of retail transactions still happen in-store, roughly 26 to 27 years after ecommerce became mainstream.
“Did ecom get rid of stores? It got rid of stores that were irrelevant or non-performing,” Kliger said. “But what it didn’t do is get rid of the stores that did it well, that had great associates who can really, really sell.”
The second example was mobile POS, which was expected to eliminate the wrap desk. It did not do that either. Instead, it gave associates more information without requiring them to leave the customer. Kliger said KWI data shows that when associates never have to leave the customer to check inventory or look up purchase history, engagement increases and sales follow.
“Mobile point of sale, like AI today, is empowering associates with so much more information, leading actually to increased sales because the customer experience is that much better,” he said.
Why Customers Still Come Into Stores
A core part of Kliger’s argument rests on why shoppers choose to visit a physical store in the first place. Customers come in for styling advice, gift recommendations, product expertise, help with returns and exchanges, and human interaction, needs that digital channels do not fully satisfy.
“If the customer just wanted a QR code or some automation, they would just shop online,” Kliger said. “I’m looking for help.”
He pointed out that the cost of digital customer acquisition exceeds in-store acquisition, and that online return rates run higher than in-store return rates. The in-store experience, Â touching a product, hearing a salesperson’s story about how it was made, seeing how other customers engage with it, continues to drive purchase decisions in ways that screens cannot replicate.
Where AI Actually Adds Value
The most practical applications Kliger described focus on removing friction at the moment of sale. When a customer asks whether a product is available in their size, what colors it comes in, or whether it can be shipped from another location, AI helps associates answer those questions faster and without interrupting the interaction.
Kliger cited Pandora as a direct example. The jewelry brand implemented an AI-powered product search in September and saw product lookups become 66% faster, with 30 seconds saved per item and item-to-cart speed doubling. The improvement addressed a specific operational challenge: Pandora does not use barcodes on displayed products, so associates previously had to look up items manually in a reference book.
“This empowers brand new associates who before did not have the knowledge of how the book was laid out with all the products to actually perform like they’re in the top 5% of that store’s staff,” Kliger said.
He also described a capability within KWI’s platform that allows associates to search for products by speaking directly to the POS. For example, asking for skin creams that do not contain a specific allergen. “Today if I want to sell you some skin cream and you tell me that you’re allergic to this particular ingredient, I don’t even think that’s possible without AI,” he said.
The Architecture Question: Unified Commerce vs. Bolt-On AI
Kliger was direct about the limits of AI layered onto fragmented systems. Without clean, unified data, AI cannot deliver the full picture it needs to be useful, and may lead to poor decisions.
“Having AI bolt on top of a POS system versus a unified commerce platform will not reach the goals that brand is looking to achieve because they’ll only have a certain amount of data,” he said.
A unified commerce architecture, by contrast, gives associates a single view of the customer, inventory, order history, returns and fulfillment options across channels. That foundation enables endless-aisle selling and smoother returns, all in one interface, without toggling between systems.
He pointed to John Hardy as an example: the brand drove a 23% sales lift on targeted campaigns because its customer data was clean and unified. The contrast, Kliger noted, is retailers who maintain separate customer records for online and in-store purchases, making it impossible to get a complete view of that customer’s history or lifetime value.
The Associate Role in 2030
Looking ahead to 2030, Kliger described a shift away from the cashier function and toward a more advisory role. Routine tasks such as paperwork, lookups and calls to other stores will be handled more efficiently by AI, freeing associates to spend more time selling and solving customer problems.
“In 2030, the associate is not a cashier,” Kliger said. “They’re a ‘clienteling’ expert, a stylist, a problem solver, an expert in the product, a brand ambassador.”
He compared the future associate to a car dealership salesperson who knows every feature of every model, except that retail involves thousands of products across different sizes, styles and customer preferences. AI, he argued, gives associates the product knowledge depth needed to have those expert conversations at scale.
Kliger also made a point about employee retention. Better tools do not just improve the customer experience, they make the job less frustrating and more effective, which reduces turnover. “People leave their companies because their job is too hard,” he said. “They’re not given the right tools.”
Measuring the Return on AI Investment
When asked how retailers should measure success with AI, Kliger framed it as an investment question, not a cost question.
Key metrics he identified include time to serve a customer, number of transactions, line items per transaction, employee retention, revenue and profitability. “AI cannot be an expense that will take more margin out of your business,” he said. “AI has to actually increase your retention, your profitability and your revenue because it’s an investment, not an expense.”
His closing message for specialty retail leaders was straightforward: “Retail is about people. AI just lets the good ones become great. And the retailers seeing the greatest returns are the ones doubling down on those people, not cutting them. Because the ones that cut them, by 2030, they probably won’t be around.”
