Retail Dive reported on October 13 that Target and Kohl's have each rolled out AI gift-recommendation tools for the holiday season, as younger shoppers increasingly treat chatbots as a discovery channel rather than a search box. Attentive data cited in the same coverage shows that 80% of Gen Z consumers used AI to help with a purchase in the past three months. That figure reframes the holiday question. The contest is no longer only about ranking in a results page; it is about being the product a chatbot names when a shopper asks what to buy.
Key Conclusions
The shift is structural rather than seasonal. When a shopper types a query into a search box, the retailer competes for a position on a list. When a shopper asks a chatbot for a recommendation, the assistant returns a short set of options and often a rationale. That compresses the consideration set before the shopper ever sees a product grid, which means the information a brand publishes becomes part of the merchandising surface. Brands that supply structured, verifiable product facts are easier to include in an answer than brands that rely on marketing language.
The second conclusion concerns the composition of the holiday audience. Gen Z and Gen Alpha shoppers arrive with different habits: shorter attention to long-form pages, higher trust in peer and assistant recommendations, and a stronger preference for products that are easy to explain.Attentive data shows 80% of Gen Z consumers used AI to help with a purchase in the past three months (Retail Dive) Retailers that build gift tools around this behavior gain a discovery advantage during the weeks when browsing volume peaks.
Where the Holiday Discovery Shift Shows Up
From ranked lists to short answers
A ranked list leaves the choice to the shopper; a short answer narrows it. The practical effect is that inclusion matters more than position. If an assistant recommends three options and a brand is not among them, no amount of on-page optimization recovers the lost visit. This is why product data quality, including specifications, materials, compatibility and availability, now behaves like a merchandising asset rather than a back-office record.
Why assistants favor verifiable product facts
Language models tend to prefer claims that can be traced. Research on generative engine optimization found that adding citations, statistics and expert context raised the likelihood of content being quoted by 30% to 40% (Retail Dive). For product pages, that translates into concrete specifications, review counts, test results and clearly dated availability information rather than adjectives. Retailers that already maintain clean catalog data are positioned well; those that do not have an unexpected remediation project.
Best Practices
Publish product facts in machine-readable form
The first practical step is to expose the attributes an assistant needs in structured fields rather than in prose. Dimensions, materials, compatibility, care instructions, warranty terms and return windows should exist as discrete values with units, not as sentences inside a description block. This is unglamorous work, but it determines whether a product can be compared at all, and assistants cannot recommend what they cannot compare.
Design gift tools around the question, not the catalog
Recommendation tools perform better when they begin with the shopper's constraint, such as a budget, a recipient and an occasion, and then narrow. Building the flow around the catalog hierarchy produces a browsing experience with extra steps. The design test is simple: a shopper should be able to reach a shortlist without knowing any product names, which is precisely the situation an assistant is built to handle.
Retailers should also track where assistant-driven traffic lands. Referrals from AI tools behave differently from search referrals: sessions are shorter, intent is higher, and the landing page is often a product detail page rather than a category page. Instrumenting that distinction is what makes the channel measurable rather than anecdotal, and it gives merchandising teams a reason to invest in the data work.
Common Mistakes
The first mistake is treating AI-driven discovery as a marketing channel and staffing it with campaign owners. The work is closer to catalog operations, because the inputs are attributes and availability rather than creative assets. The second mistake is optimizing for volume of content instead of clarity of facts. Assistants do not reward length; they reward specificity that can be matched against a question.
The third mistake is assuming the shift applies only to young shoppers. Gen Z leads adoption, but the same behavior spreads upward as assistants become default interfaces on phones and in messaging apps. The fourth mistake is measuring the channel by last-click attribution alone. When an assistant shapes a shortlist, its influence is real even when the final purchase is attributed to a later direct visit.
Reading the New Discovery Funnel
The funnel now has an invisible stage. Before a shopper reaches a retailer's site, an assistant may already have reduced the field, and that reduction leaves no session in the analytics platform. The only durable response is to make the brand legible to the assistant, which means maintaining structured data, dated availability and traceable claims. This is a different investment profile from paid acquisition, and it compounds rather than resets each season.
Holiday planning should reflect that. Retailers that spread promotions across more weeks and publish clear availability information give assistants something stable to reference.NRF publishes holiday sales forecasts and consumer spending expectations that frame the winter shopping season (NRF) In a season where the consideration set is set earlier, the retailers that are described accurately are the ones that get asked for.
Summary
Target and Kohl's AI gift tools, together with the finding that 80% of Gen Z shoppers used AI for a purchase decision in three months, mark a change in how holiday discovery works. The consideration set is narrowing before a product grid is ever rendered, and assistants favor facts that can be traced. Product data quality has become a merchandising input rather than a back-office concern.
The practical agenda is to publish product attributes as structured values, design recommendation flows around shopper constraints, and instrument assistant-driven referrals so the channel is measurable. These steps are unglamorous and they compound. In a season where the shortlist forms early, being legible is the whole game.
Data Sources
FAQ
What did Target and Kohl's launch for the holidays?
A: Both retailers rolled out AI gift-recommendation tools aimed at younger shoppers who increasingly use chatbots as a product discovery channel rather than a traditional search box.
How many Gen Z shoppers used AI for purchases?
A: Attentive data cited by Retail Dive shows that 80% of Gen Z consumers used AI to help with a purchase in the past three months, making assistants a mainstream discovery path for the group.
Why does structured product data matter for AI discovery?
A: Assistants can only recommend what they can compare. Attributes stored as discrete values with units, such as dimensions, materials and compatibility, are easier to match against a shopper's question than the same facts buried in descriptive prose.
How should retailers measure assistant-driven traffic?
A: By segmenting referrals from AI tools separately from search referrals and tracking their landing pages. Sessions are typically shorter with higher intent, and the landing page is often a product detail page rather than a category page.
Is this shift limited to younger shoppers?
A: No. Gen Z leads adoption today, but the behavior spreads as assistants become default interfaces on phones and in messaging apps, which gradually pulls older cohorts into the same discovery pattern.
