Our Retail expert, Zainab Hussain, is an e-commerce strategist with extensive experience in customer engagement and operations management. In this discussion, we explore the shift from keyword-based search to intent-aware systems, the mechanics of modeling shopper behavior at scale, and the practical integration of AI tools for store associates. We also dive into the “super agent” concept and how brands can balance investment between traditional search traffic and the high-conversion potential of agentic commerce.
Traditional keyword-based searches often return null results for lifestyle queries like “what to wear to a music festival.” How does transitioning to intent-aware search change the discovery process, and what specific metrics should merchants monitor to ensure the system correctly interprets these complex shopper needs?
The shift to intent-aware search moves us away from rigid text matching and toward a system that understands human context. When a shopper asks what to wear to Coachella, an intent-aware system—driven by tools like CommerceGPT—knows to surface sundresses and boots rather than delivering a frustrating “no results found” message. For the merchant, the discovery process becomes less about manual tagging and more about letting the Large Language Model infer meaning from the product catalog. To ensure this is working, merchants must closely monitor “null result” rates and conversion lift on long-tail queries. By looking at how about 20 early adopters are already seeing benefits, it is clear that tracking the journey from a vague lifestyle query to a completed purchase is the best way to validate that the AI truly understands the shopper’s heart.
Some systems now simulate customer queries tens of thousands of times to mirror the search depth of major global marketplaces. How does this modeling process function during a sudden trend shift, and what steps are necessary to ensure the AI responds faster than a human merchant could?
This modeling process works by taking a merchant’s existing catalog and search history and then simulating thousands or even tens of thousands of potential interactions. This massive scale of simulation allows a mid-sized retailer to have the same depth of data and foresight that a giant like Amazon possesses. When a sudden trend emerges on social media, the AI detects these shifts in real-time by recognizing new patterns in natural language and intent expressions. Instead of a human merchant having to manually listen to social channels and update the catalog, the system automatically recalibrates to provide relevant results. To keep this speed advantage, retailers must ensure their product data is clean and that the AI has permission to adjust search rankings dynamically without waiting for a manual weekly refresh.
When technical product data is integrated into search tools, store associates can use the technology as an on-the-floor training and sales resource. What are the practical steps for deploying this technology across physical locations, and how does it change the interaction between staff and customers?
Deploying this in-store begins with giving associates mobile access to the same contextual search tools used on the website. For a merchant with a highly technical catalog, the practical first step is ensuring the AI can handle deep-dive product details that a human might not memorize. This changes the interaction from a simple “let me check the back” to a high-level consultation where the staff member can speak expertly about complex specifications. We’ve seen that when associates use these tools, they become better trained on the fly, connecting shoppers to the exact technical specs they need. It builds a deeper level of trust because the associate is no longer just a salesperson; they are an empowered expert supported by real-time data.
The “super agent” concept involves orchestrating multiple specialized AI agents to handle tasks like comparison shopping or order management. How can brands prevent customer friction when managing these multi-agent interactions, and what strategies keep the experience focused on convenience rather than technical complexity?
The key to preventing friction is ensuring the customer only ever feels like they are talking to one cohesive entity, even if multiple specialized agents are working in the background. Shoppers likely won’t tolerate having to manage five different agents for one purchase, so a “super agent” acts as a conductor to orchestrate information seamlessly. This approach keeps the focus on the outcome—like finding a margarita machine that is easy to clean for two dozen people—rather than the technical steps of searching. Brands should focus on a “one-interface” strategy where the orchestration happens behind the scenes. This maintains the convenience we all grew accustomed to during the pandemic, where the goal is simply how fast a solution can reach the doorstep.
Shoppers arriving from AI-driven sources often show a high “ready to buy” intent but still represent only a portion of total web traffic. How should retailers balance their investments between traditional search and these new agentic sources, and what anecdotes illustrate the difference in these shopping behaviors?
Retailers are currently facing a balancing act because, while agentic traffic is growing, the majority of sales for the 78 Salesforce-using retailers in the Top 2000 still come from traditional search and paid media. The difference in behavior is striking: a traditional shopper might spend ten minutes filtering by size, color, and fabric, whereas an AI-referred shopper arrives with their mind already made up. One anecdote involves a shopper who bypasses all traditional site navigation because an LLM like ChatGPT has already narrowed down their choice to a single item. Because these shoppers are so “ready to buy,” the conversion rates are higher, which justifies a focused investment in API readiness for agents. However, merchants cannot abandon traditional SEO, as $192.60 billion in combined sales for these top retailers still relies heavily on the legacy search funnel.
What is your forecast for agentic commerce?
I believe agentic commerce will fundamentally transform the entire customer journey by moving us into an era of “anticipatory retail.” We are moving toward a future where agents don’t just find products, but actively manage orders, handle complex customer service returns, and even coordinate with other agents for comparison shopping across different brand sites. While it is currently a nascent area, the demand for extreme convenience will push these tools to handle the “boring” parts of shopping, like filtering and technical comparisons. My forecast is that within the next few years, the success of a brand will be measured not just by its website traffic, but by how well its data can be “read” and prioritized by these autonomous AI agents.
