Imagine an online shopper getting ready to host a tailgate.

Just a few years ago, they’d have to think through every item they needed and search for them one-by-one — typing “burgers,” then “buns,” “cooler,” “trash bags,” “solo cups,” “speakers” and so on into the retail search bar.

If they tried something like “Help me plan a tailgate,” the search engine would’ve laughed at them — if search engines were sentient, that is.

Today, they’re starting to feel a little more human. Shoppers can type in open-ended requests like that on many retail sites and get personalized, cross-category suggestions in return, much like they would from a helpful in-store associate.

This behavior is only going to grow. Retailers need to be ready.

The age of AI agents

Conversational searches aren’t niche anymore: They’re quickly becoming a natural extension of how people use AI in their daily lives.

In fact, data from Constructor and Shopify’s State of eCommerce report shows that nearly two-thirds of shoppers have used generative AI tools like ChatGPT in their everyday lives — a staggering 121% increase from just two years ago. And that number is only going to rise.

Unsurprisingly, this shift is reshaping how people want to shop:

60% say they’d let an AI shopping assistant guide them when they’re unsure what to buy.

And nearly 4 in 10 (38%) have already tried agentic shopping tools, like Amazon’s Rufus or Walmart’s Sparky, on retail sites. For Gen Z, it’s even higher (51%).

Among those who’ve used AI agent features on eCommerce sites, half (50%) say the experiences are “always” or “often” helpful, and 86% say they’re helpful at least some of the time.

As shoppers shift from searching to asking, retailers that deploy AI agents delivering contextual, cross-category guidance will win loyalty, drive higher conversions, and define the next era of ecommerce.

Small user base (so far!) but mighty results

Right now, only a subset of shoppers uses AI agents on retail sites. But as the data above shows, that number will multiply fast. And even in these early days, their business impact is hard to ignore.

One retailer, 200+% gains: I was recently speaking with a C-level exec at a leading specialty retailer who’s thrilled with how they’re using on-page AI agents. While only a small percentage of traffic engages with the agents today, those shoppers are converting at far higher rates — driving 200%+ more add-to-cart activity and 200%+ more revenue compared to non-users.

Black Friday proved the point: Constructor’s data from Black Friday 2025 highlighted the growing impact of AI shopping agents. On sites that use them, more than 10% of Black Friday/Cyber Monday revenue came from shoppers who engaged with an AI assistant.

Global trend, fast growth: Salesforce also recently reported that AI and agents influenced 20% of all global orders during Cyber Week, and retailers who used branded agents saw 32% faster sales growth. It’s another clear sign that AI agents are becoming a meaningful driver of eCommerce performance.

Looking ahead

Retailers are moving past the novelty phase of AI agents. What started as experimentation is quickly becoming more intentional, driven by results, not hype.

Here’s what we can expect in the year ahead:

1. Goodbye shiny, new object; hello ROI. Retailers will move beyond what’s flashy to what works. Spurred on by the wins above, and a clearer sense of what’s possible, they’ll prioritize applications with measurable results, like increased conversions, fewer returns, and faster resolution of shopper questions. We’ll continue to see value-driven implementations across the shopper journey: in product discovery, customer service, returns, post-purchase support and more. As retailers evaluate these agentic commerce use cases, the question won’t be “Does this look cool?” but “Does this add value to shoppers and the business?” And “Does it support how shoppers want to (and will want to) shop?”

2. Sharper use cases. When it comes to helping shoppers discover items they need and love, agents will increasingly guide them at specific and key moments in the shopper journey. This includes early-stage research (queries like: “I’m starting rock-climbing — what gear do I need?”) and at the point of decision on product detail pages (e.g., “Will this harness fit over winter layers?”). Guided assistance is critical during these points: reducing friction, increasing shoppers’ confidence in their purchases, and lowering the likelihood of returns. It’s a win-win for shoppers and retailers alike.

3. The next frontier: Agent-to-agent handoffs. As more product discovery starts offsite — in AI platforms like ChatGPT, Perplexity and other assistants — retailers will look for ways to meet shoppers where they are, and earlier in their journeys. Agent-to-agent handoffs will make that possible: One assistant helps a shopper get started offsite, then passes the context to the retailer’s onsite agent to pick things up where the shopper left off. The result? A smooth, connected experience, from first question to final purchase.

Cooking up grade A experiences

As product discovery migrates to AI, retailers need to bring AI-grade discovery to their own properties — not just to compete, but to truly connect with how people shop today.

Much like the tailgate shopper at the start, consumers aren’t looking to piece things together one search at a time. They want help planning the whole spread.

With AI agents, retailers can deliver that kind of intelligent guidance and serve up the kind of shopping experiences that leave everyone satisfied.