Insights from our retail roundtable, part 2
At our recent private roundtable with retail leaders, the conversation kept returning to a practical question: where is AI creating the most tangible impact right now?
When service contact data becomes a mirror
While discovery tools attract attention, some of the clearest results have emerged in customer service. Retailers sitting on thousands of customer service conversations are finding that AI classification models reveal something unexpected. What starts as an efficiency play, reducing call volume, shortening handle times, quickly surfaces something more valuable: a clear picture of where the operation is actually breaking down.
The patterns tend to be consistent. Recurring contact reasons that point to broken processes, self-service journeys that fall apart at critical moments, and product and delivery issues surfacing in the same places again and again.
For retailers willing to act on that signal, the opportunity isn't just to deflect contacts. It's to eliminate the reasons customers needed to reach out in the first place.
But the more provocative questions emerged from there: If conversational AI can identify friction in service, can it also predict opportunity upstream? Could recurring product questions signal assortment gaps? Could language patterns hint at emerging trends before sales data confirms them? Could sentiment shifts inform merchandising strategy?
The ambition is to move from reactive automation to predictive intelligence. Most organizations are still in the diagnostic phase, but the direction is clear.
And this is where the agentic commerce thread becomes impossible to ignore. As AI agents begin acting on behalf of customers, they won't tolerate a broken self-service journey or a misrouted return claim. They'll simply go elsewhere. Retailers who have used service contact data to eliminate friction will have operations reliable enough to serve the next generation of buyers, one that may not be human at all.
Beyond checkout: the automation horizon
While much of the conversation focused on discovery, agentic commerce extends further. Participants described early explorations into automated return-label generation, QR-based return processing, and conversational assistance for delivery claims.
Each use case removes a small piece of friction. Individually, these features appear incremental. Collectively, they reshape expectations. When friction disappears in one interaction, tolerance for friction elsewhere declines.
Retailers aren't yet witnessing fully autonomous shopping agents transacting at scale. But referral patterns from conversational systems are already changing attribution models. Understanding less about the path doesn't necessarily mean understanding less about the customer, but it does require new measurement frameworks.
Intent, friction, and the human layer
Across multiple categories, one theme surfaced repeatedly: intent matters. Some leaders described deliberately leaving good friction in the customer journey – pauses or decision points that allow customers to experience discovery, delight, and surprise.
Conversational AI isn't just a shortcut. When used thoughtfully, it can preserve and even enhance the emotional connection that distinguishes one brand from another.
