Ulta Beauty had a strong quarter. Net sales hit $3.16 billion, up 11.1% year over year. Comparable sales grew 5.3%.
But the part worth paying attention to is the split. Ecommerce comps grew in the mid-teens, while physical stores crept up by low single digits. That is a strategic fork in the road.
Enter the AI Shopping Assistant
In April 2026, Ulta launched what it is calling Ulta AI, a conversational shopping assistant built on Google’s Gemini Enterprise platform. It lives on Ulta.com and inside the app, and it does roughly what you would expect: you type in something like “a foundation for oily skin under $40” and it finds something instead of making you scroll through 2,000 results.
CEO Kecia Steelman called the early results “promising.” Which, in corporate speak, means it is probably working well enough to keep funding.
Bigger Baskets, Not More Shoppers
The growth was not coming from new customers walking through the door. Average ticket rose 3.7% while transaction volume grew just 1.6%.
Roughly 70% of Ulta’s comparable sales growth came from people spending more per visit, not more people visiting. An AI assistant that recommends the right serum to go alongside the foundation you already put in your cart is a margin instrument.
How They Paid for It Without Blowing Up Their Margins
Ulta’s gross margin expanded by about 90 basis points to 40.1%. Then they let SG&A costs rise by about 80 basis points to 25.7% of sales. Operating margin landed roughly flat at 14.2%.
That is not an accident. Ulta took the margin it gained on the product side and reinvested it into technology and distribution, rather than letting it fall to the bottom line. It is a clean playbook for funding an experiment without spooking investors.
Not Just the Chatbot
Ulta did not drop one AI tool and call it a transformation. In the same quarter it launched a TikTok Shop storefront, a single April livestream pulled more than five million impressions. It also rolled out Uber Eats delivery across 1,500-plus stores and expanded its Klarna buy-now-pay-later integration.
The pattern is deliberate: meet the customer wherever they are already spending time. That is what “distributed commerce” actually looks like in practice.
The Target Breakup
Here is the context that makes Ulta’s digital push look less like ambition and more like necessity.
In August 2025, Ulta and Target announced they would not renew their in-store partnership. Ulta had an estimated 600 shop-in-shop locations inside Target stores, representing nearly a third of Target’s 1,981 U.S. stores. All of those close by August 2026.
TD Cowen analysts pointed to staffing, loyalty program synergies, shrinkage, and return on invested capital as the likely factors behind the split. Which is analyst-speak for: it stopped making financial sense for both sides.
That is around 600 physical touchpoints vanishing from the map within the same year Ulta is posting mid-teen ecommerce comps and launching a Google-powered AI assistant. The timing is not coincidental. Ulta is not just growing online. It is rebuilding how it reaches customers after losing one of its biggest retail partnerships.
Does AI Actually Sell More Stuff? Retailers Want You to Think So
Look, every major retailer in 2026 is either piloting an AI shopping tool or about to announce one at a conference. The word “conversational commerce” is showing up in earnings calls the way “omnichannel” did a decade ago: sincerely, repeatedly, and occasionally without much substance behind it.
The honest reality is that Ulta cannot fully separate the AI assistant’s contribution from the broader investment in loyalty, new channels, and merchandising. Mid-teen ecommerce growth and a 3.7% ticket lift are consistent with a working AI discovery layer. They do not prove it.
What This Means for Ecommerce Businesses
The Ulta story is useful for ecommerce operators beyond the beauty category. Here are a few things to take from it.
If your traffic is flat and your conversion rate has plateaued, your growth is almost certainly hiding in basket size. Recommendations, bundling, and AI-assisted discovery all serve the same function: they help the customer who already showed up spend closer to what they actually intended to spend.
The “distributed commerce” pattern matters too. TikTok Shop, Uber Eats, Klarna. None of those are revolutionary on their own. Together they describe a company that has stopped relying on the customer coming to its own front door and started showing up in every channel where purchase decisions get made.
And the margin discipline is the part most worth copying. Funding innovation from gross margin improvement, while protecting operating margin as the floor, is the adult version of “we are investing in growth.”
Our Take
Retailers Are Selling You on AI Before They Have Proof It Works
Ulta had a genuinely good quarter and the AI assistant is a real part of the story, not just a press release prop. The more interesting thing is that it is happening against the backdrop of walking away from 600 Target locations. This is a company consciously shifting its distribution model under the surface of a strong earnings report.
Whether conversational AI becomes a durable ecommerce lever or a 2026 trend that quietly disappears into the FAQ section of every website, we will know more by Q3. The brands worth watching are not the ones who announce AI features. They are the ones who publish the holdout test results.
For now, Ulta is doing the right things in roughly the right order. And its chatbot is probably selling more mascara than most people’s entire marketing teams.
