Ulta AI chatbot pilot lifts ecommerce past store growth

What's happening

Ulta Beauty posted Q1 fiscal 2026 net sales of $3.2 billion on June 2, 2026, up 11.1% year over year with comparable sales up 5.3%. On the same earnings call, CFO Christopher DelOrefice told analysts the ecommerce channel delivered "mid-teen sales growth", far outpacing comparable physical store growth, and the company shared early results from its AI shopping assistant pilot. Digital Commerce 360 covered the AI numbers on June 5; the earnings press release sits on the SEC EDGAR filing for Ulta.

The headline that traveled fastest framed it bluntly: Ulta has built a chatbot that is moving more inventory than several of its physical stores. The underlying story is more nuanced, but the direction is clear. Ulta is one of the first US specialty retailers to publicly attach revenue growth numbers to a generative-AI shopping surface, and it landed two weeks before Prime Day.

Why this matters beyond Ulta

Beauty is a category Amazon, TikTok Shop, and Sephora all want a bigger share of. Ulta publicly tying generative AI to double-digit ecommerce growth gives every other retailer cover to invest in conversational discovery surfaces, which means Amazon Rufus is no longer the only place sellers need to think about agentic discovery. For brands selling on Amazon plus a DTC store, plus marketplaces like Walmart and TikTok Shop, the discovery layer is fragmenting faster than the ad layer.

The second-order effect is on attribution. When part of the customer journey moves into a chatbot, last-click reporting on the marketplace side gets noisier. Brands that already blend Amazon spend with off-Amazon channels in a single P&L see this earlier than brands relying on Seller Central reports in isolation.

What multi-channel sellers should do this quarter

1. 1.

   Track Amazon Rufus share alongside off-Amazon AI surfaces

   Rufus, ChatGPT Shopping, Perplexity, and now retailer-native chatbots are converging on the same shopper question. Tag any off-Amazon traffic referred from an AI surface as its own channel in your analytics so the mix is measurable in 90 days.

2. 2.

   Audit beauty and personal care ASINs for discovery readiness

   Conversational surfaces lean hard on structured attributes — shade, finish, ingredient, skin type. If your Seller Central listings still rely on bullet copy to convey those facts, the AI layer cannot quote you. Fill the attribute fields on your top revenue ASINs first.

3. 3.

   Blend Amazon and off-Amazon revenue in one P&L view

   A retailer-by-retailer view masks the moment when one channel is silently subsidising another. Pull Amazon, Shopify, and other marketplace sales into a single P&L so a shift away from store traffic shows up in margin, not just channel-mix slides.

4. 4.

   Hold a per-SKU margin floor before discounting for retailer events

   As more retailers run AI-personalized promos, the temptation to match every deal grows. A clear floor per SKU, set against contribution margin in custom analytics, keeps off-Amazon promos from eroding the same SKUs that fund Prime Day on Amazon.

How Nova helps

Nova reconciles 40+ Amazon fee types at the SKU level and lets brands blend Amazon performance with off-Amazon channels through the Data API. As discovery splinters across Rufus, retailer chatbots, and AI search, the brands that survive the transition are the ones who can already see contribution margin per SKU, per channel, on the same day. Coverage spans the 21 Amazon marketplaces Nova supports, with FBM included on the same dashboards as FBA.