Is agentic shopping the next big thing in beauty? Sephora and Ulta are betting yes

[MUSIC] Hello and welcome to the Glock P Beauty Podcast, or show about the business of beauty and wellness. I'm Sarah Sprick-Finer, and I'm Lexi Lebsack. On May 13th, Amazon announced that it is bringing its AI assistant Alexa Plus directly into the site shopping experience, replacing its Rufus AI. In an interview with the Virg's Vice President of Alexa and Echo, Daniel Rouse, said that the new Alexa for shopping will be more deeply integrated, more capable, and available everywhere. As we've talked about on this show before, Amazon is trying to further cement their place as a go-to for beauty shoppers, competing with the Foren Ulta. Amazon may have a leg up with its own proprietary LLM that powers Alexa Plus, but the Foren Ulta are not getting left in the dust. They both launched AI partnerships in the past two months. The Foren with Open AI at the end of March, and Ulta Beauty with Google Gemini at the end of April. We wanted to understand why beauty's key retailers are rushing into AI. To talk about the Foren Ulta's AI strategies, how they differ, and where Amazon fits in, we are joined by Glossy Beauty Senior reporter Emily Jensen. Hi, Emily. Hi, Sarah. Hi, Lexi. Hello. So, to kick us off, Emily, tell us about why Ulta and Sephora are re-thing to implement AI into their digital forefront. Well, from what I'm hearing from people who are more AI experts than I am, is it's kind of exactly what you were saying before. They are seeing the competition from Amazon. Amazon, as you said, has been building out for years its own proprietary artificial intelligence technology, and that's just something that very few companies have the infrastructure resources to build out on their own. So, it makes more sense at this point for a Sephora and Ulta type company to then outsource to that infrastructure. So, as you said, Sephora has gone with Open AI. They have announced what is sort of an integration between the Sephora app and the ChatGPT app. This infrastructure kind of already existed. ChatGPT has been announcing these partners with other big companies like Spotify or Zillow or also part of this partnership. Essentially, it's like if you're a ChatGPT user and a Sephora user or loyalty member, you can link your Sephora app to your ChatGPT app. So, it can kind of help you shop your Sephora picks. And in theory, in the long term, you'll be able to check out directly in the ChatGPT app. While getting your Sephora loyalty points and whatnot. Ulta, as you mentioned, has gone for the other big AI company with Google. And announcing more of an integration is a bit different from my understand. You'll be able to shop on Ulta.com with their AI assistant, which will now be powered by the Google Gemini technology. Because again, rather than them trying to build out their own AI bot, that's just not going to really have the sort of capabilities of the big tech juggernauts. They're just outsourcing that and bringing it in to their own sites. And then down the line, from my understand, hasn't maybe been rolled out necessarily yet. You'll be able to shop for Ulta Beauty on your Google Gemini page as well. So again, ultimately, a lot of this is just about bringing, I guess, more beauty companies, like seeing the writing on the wall, seeing that people are using things like ChatGPT or Google Gemini to kind of discover beauty products to learn about them. And then wanting to then insert themselves into this pipeline, hopefully so that customers are maybe using ChatGPT, but not walking away from shopping at Sephora, more Ulta in the long term. Emily, you published a story a little while back that had some really interesting stats around the sort of why brands, why retailers really want their products to appear on AI platforms, like ChatGPT or Google Gemini. Can you share some of those with us? What did you learn? Totally. I mean, this is a thing yet. It's like the growth in how people are using their AI chatbots is really exploded in the past couple of years. And it's still, I imagine, changing quite a lot as consumer behavior kind of adapts to these new platforms, but just for some stats to give a sense of how many people are using them. This was data that I'd pull it from Mielsen IQ. They stated that customers conduct over one billion beauty related searches per week on ChatGPT. Holy moly, that's wild. A billion a week. I know. This also comes from an April, 2026 report from a company called Tenuity, which said that more than a third of beauty shoppers are using AI to research or purchase products. That proportion goes up to more than half of beauty shoppers when we look at Gen Z specifically. So again, like a good chunk of beauty consumers are moving in this direction. And it's kind of, it's interesting to think about too. It's still such a polarized world of AI. I'm curious like how you guys see this in your life, but there are people who, like I know, for example, who are really early adopters of ChatGPT and really use it heavily. And then some people who are actively resistant to it and like do not want to go on any AI platform at all. And there's some people who are maybe in the middle who are like, maybe I use it for this thing here and there, but I'm not turning to chat every time I want to know like, okay, what's the best, you know, blush for my skin tone or what's the best like tinted moisturizer that I should try? Oh, 100% I know people that literally don't use Google anymore and only use AI. And I know people that think it's morally reprehensible to ask it one single question. I'm very curious if you guys, I'm assuming that you guys haven't really used it for beauty in your own lives, but if you have and also if you know of anyone in your life who has. I know quite a few people who are using Chat sort of every day for every question, ChatGPT. I know quite a few people like that. I'm more of a textbook later, doctor. I did try all of these things of course going into this episode. And I found they were all like somewhat similar, but had some nuances of course between the different sort of providers and platforms. Yeah, I don't know. I'm sort of a late adopter, but I'm coming around to it. I feel like every time I Google something, Google like kicks me to Gemini, like they're really trying to convert me, which I don't necessarily mind. I'm sure I'll be using Gemini eventually. Emily, do you use a do you use Chat or do you use any of these AI platforms? No, I think I'm similar to you that way where at this point, yeah, I wouldn't call myself an early adopter of AI or any of these new tech things. And I do find it sometimes like frustrating just on any online experience now that like AI is being sort of forced at you and like cluttering the experience in a way that I just personally don't find super helpful. But I did tinker around with it. Again, just as we're writing about it so much, it's like helpful to try and understand like what are we actually talking about? Like what is the real experience there? But so thing too with Chat GBT or any of these AI bots that's interesting is I was talking to Megana Dar for this piece I wrote about recently about AI. She is a sort of technology advisor. She used to work at meta overseeing their Instagram shopping and she you know very she lives in San Francisco. She's very openly like bullish kind of pro AI. And she was saying the thing with Chat GBT is that because she uses it frequently, Chat GBT like knows kind of who she is like it knows what she looks like. It knows what she likes. It knows it knows maybe a bit more about her purchasing history. So if someone like her her experience using Chat GBT is probably so much more maybe elevated and kind of richer than me or anyone who's kind of popping on it for the first time and like trying to figure it out. Like there is kind of a hurdle or a learning curve of like you have to train your chat bot to understand you and your preferences, which some people again don't want to do. But it is I am curious down the line to your point Lexi. If eventually well we just not be able to avoid using these AI bots as they do kind of get put in front of you more and more and more when you're trying to do a Google search. There were some people back and I'm old enough to remember like people back in the day being resistant to using Google for researching like you know for school and that kind of purpose. So maybe AI will go the same way. Yeah like go look it up in a book kind of vibes. I haven't heard since I was a child. But yeah I mean it's like there were late adopters back then too and of course now we're now we're all using Google like crazy. I'm curious as to Emily and Sarah both of you guys when you used these things what did you sort of think what which one do you think is best if you had to if you had to say well I find it still and again part of this maybe my own my own chat vocabulary is not well developed but I find it a bit clunky to use. I mean with the chat GPT thing it's like with Sephora there's like this sort of main chat interface where you know I played around with it just to see like okay if I ask it to recommend me some lipsticks what is it going to pull up for me I asked it to recommend some Sephora lipsticks specifically for like my skin tone what do you say I should buy. I mean you know it kind of gives me an assortment of brands and products that I again I'm like I already know about all of these brands and products this isn't really telling me anything new. There was some I don't know if this is an issue on chats end or like Sephora's and there was this like technical glitch where it was recommending Charlotte Tilbury lipstick, but the actual image that it pulled up was a merit lipstick. I'm trusting. I know. So, but I don't know what that's about. I'm sure there's tons of little glitches on the back end with that sort of thing. But if you're trying to actually get to this new integration, you have to like create a chat to be T-account and actually manually link your Sephora app into your chat account, which again for some people is maybe very natural and they're very happy to do that. For me, it's a bit like I don't really see what this is adding to the experience quite yet. But again, that goes down to I'm also not shopping on Sephora, a ton. If I was shopping on Sephora, a ton and I had a really developed profile there. And I was a user of like, well, I want to retain all of that profile and points and like, you know, cart history, built up there. And now I want to integrate. I want my chat to be T-to-know about that history. So, it can better help me shop. Maybe that would be more helpful. Yeah. I don't have a huge history with Altersso for although I do shop on their sites, but I do order some things on Amazon and I shop at Whole Foods. So in one of the responses I got playing with the Alexa AI chatbot, it did sort of say, like, well, according to your shopping habits, you like to buy organic food. And so what about these supplements, which I'm not sure I necessarily saw the connection between the supplements and the fact that I bought organic food at Whole Foods, but it was certainly trying. It was certainly like reaching back as best it could to try to pull some information and make it interesting to me. Yeah. So have you tried these at what sort of your takes on them? Yeah. So I asked my friends if they had used AI for beauty because I recognize that I'm not a traditional beauty consumer. And one of my best friends that she had, she was very cute and bashful about actually putting it having wasted water for this, but she had used it for something that immediately clicked for me as, okay, this makes so much sense for color matching. And I said, yeah. So we had a conversation about it. And I said, you know, like, oh, like I'm so skeptical and resistant to give it any image of me. And she said she'd cut off her eyes. And I was like, oh smart, you know, my friends are smart. And so then I did that. And I did it specifically. I, you know, I took a snap to quick selfie, cut my eyes out of the picture and he identified like super bright and a faint part of my face. And I asked it to color match me in my current favorite foundation that I wear all the time. But I'm not even sure I'm wearing my right shade, which is the Mary Phillips MPH's Listkin. And I asked Gemini and Chat GPT. I sent it the same photo and said what would be my shade in this foundation. And they both said the same thing. But I in terms of color matching, but it was so informative to me because even though that's a very specific use case, it makes so much sense that, you know, the past 10, 15 years editorial and I've written many of these stories before I got to Glossy, it was like, you know, a SEO arms race of like the 15 best foundations for, you know, mature skin for dewey skin, whatever. And this is such an obvious one up to that when you can just go and put in a picture of yourself and some people don't care. They'll just put a full picture of their face and get, you know, you could say I have acne prone skin that's dry on my cheek than whatever and get these super custom tailored recommendations. And that was just so fascinating to me. In this case, it didn't, I mean, it doesn't, the retailer part wasn't necessarily part of the equation because MPH is still exclusive to Sephora. But it does make so much sense to me to bring it full circle that, you know, they need Sephora and Ulton need a new way to compete with Amazon, you know, as we were speaking out on the last episode, the three of us did together because so many brands are now available there too. I am really curious to in the long term is interesting what you said that they both recommended the same thing like Gemini and Chatchy PT because I'm curious in the long term like how much do these chat bots look like the each other or if they're delivering up different information? Like when I was reporting on this as well, this was something that I talked to Front Row, which is an e-commerce agency that, you know, they have a lot of expertise around Amazon and kind of Google, Gemini and Chatchy T and whatnot. And Jason Wheeling-Man, who is their SVP of Marketplace Performance at this agency, had told me that Amazon does not not allow other chat bots to scrape its data, but Amazon scrapes other, you know, other sources for it. Yeah, which is very interesting of like, okay, so like I'm sure of course they all want to know what the other kind of bots are doing and do they want to be aligned in the long term or do they want to offer different results? You know, we're going to start seeing as well more advertisements come into these platforms, open AIs, I think CEO had gotten a bit of pushback for famously being resistant to ads and now recently they've announced that they are integrating ads into chat. So it's interesting though, but I think to your point Sarah about yeah, like the hesitation of uploading a photo, I totally feel that of like, why do I, even though I'm sure it can already pull a million photos of me if I wanted to. But a lot of the the gold mine for Sephora and Ulta and Amazon too and any kind of retailer working with AI in the long run is the data. It's not even just about converting shoppers and getting them to actually like check out and buy more potentially with AI, although that is certainly a part of it in the long term, but the real value is like they can have in theory so much more data now on like, how is the person starting their shopping journey? Where are they starting from? Where are they looking for? You know, what steps do they take to then select this product and then actually check out and having that all in one window on your chat would be very, very, very valuable for these companies versus maybe what happens more so now where you might watch a TikTok, you might read a Reddit post, maybe you then go in and store and try things on in person and then you go check out maybe on Amazon later on and all of that kind of data is very fragmented of that customer journey. No, it makes so much sense that it's a play not just for the sales like it seems like maybe initially Emily from your reporting. Great if they can capture some sales excellent, but maybe not even the primary goal. It's a goal for sure and it seems like maybe it's something that they're investing in because they know that like in the future maybe more and more sales will come from AI chatbot driven conversations, but in the short term it's more of an investment in data. Totally and also this goes back to the advertising thing of you know if you are integrating more ads onto the chat experience the whole name of the game is now just keeping people's eyeballs on the window longer so that you can sell more ads. And if we have more information on like what people are looking for would actually keep them like on that window for longer then we can sell them more ads. So we just need to have again just more knowledge, more understanding of say like oh now we have a database potentially of like these 40 million Sephora loyalty members and kind of like what they're shopping for and what they're looking for import on to our platform we can better understand how to target them with this or that brand or you know or whatever maybe. I wonder if you know the brands will start to be privy to some of this data it's like I wonder if Ulta is going to be sort of slowly dispersing some of this out to the brands that are sort of connected to this or Sephora will be while I was shopping on Sephora I kept asking it questions and I felt like I kept getting the same products like I kept getting this one Sephora collection foot mask like it would it found a way to try to get me to look at this foot mask several times their several different queries it's very random and I also found that both Ulta and Sephora were pushing a lot of rare beauty products so that was also up in the top one whether it was you know a lip color or or a lotion or something but they were also pushing that very hard so I'm curious and I know we don't know the answer to this but I'm curious as to sort of where the brands will sort of benefit from this or how the brands can you know potentially you know elbow their way into the top of the search results here and and sort of what those partnerships will end up looking like you didn't find anything like that Emily that that that's like very very later on right but I think it's a really good point actually it was something I was talking with them again with Megana Megana Dara this sort of AI expert about is this integration between the tech companies and like the beauty companies who probably are speaking a very different language and it's sort of curious like what's going on in the backend that yes someone is trying to force this like foot mask to you and how was how are the products being merchandise like what is the chatbot think is like a good experience versus what a more knowledgeable you know native kind of beauty company might think and Megana mentioned a funny kind of anecdote that she used to work at eBay back in the day and they were trying to get more like apparel brands comfortable with like being on eBay or whatever, I'm paraphrasing her story, but there was this, their algorithm had merchandised like a Nike product next to an Adidas product on the page. And sports where apparel people will know, like that's a huge no-no to have your Nike socks, next to your Adidas shoes or whatever. It may be, and she was like, we got a very angry call from Nike of like, how did this? And up here, because again, like the tech person who's maybe designed that algorithm or whatever, they don't know that. They don't know that you're, that's a no-no to put these two brands next to each other. They're probably just looking and they're like, oh, they go together 'cause they're both sportswear apparel brands. I don't know what the equivalent would be. Maybe you're like, rare beauty next to your road, you know, lip liner or something. Yeah, we're like an Ilya and a safe foundation right next to each other. Right, yeah. So it is curious to see, like what brands kind, I'm sure there's some person, maybe any of those brands whose job is now to go on these chats and like look at how their products are being presented. Because this is also something that I've been told from experts is that this is a very scary kind of time for brands 'cause they're not owning the sort of brand experience of like how their products are being shown or presented or labeled versus on their sites or their Instagram where they can have a very controlled sort of environment of what their branding message is. Yeah. It'll be interesting to see how there's sort of tactics evolve. Like I know when I was shopping on Ulta versus Sephora, it was, I mean, similar obviously, they're trying to offer you products or trying to explain why. But I found that Sephora's was way more concise that it was very succinct, conversational. They did not give me any follow up it. They, oh my gosh, I'm calling the AI, I'm so sorry. It didn't give me any follow up questions. It kept all of its responses very, very short and to only three or four products where I would go to Ulta Beauty and I would get so many products. It would give me 11, 12 products broken down into different categories. And then when I went to ChatGPT, it got even more expansive where it would give me 15 products with these huge chunky sections of copy around sort of the various finishes and the various undertones. And very quickly, I can understand why someone would say, oh wow, that's too much information. But maybe Sephora isn't enough information if you don't try to grab them with another question, like who's this for? And what else are you using or whatever the follow up question is? But I'm sure we're gonna have a lot of findings soon. - It's interesting to me, just how much, it's like inescapable looking at AI now and it's kind of fascinating to me, just seeing again, for these companies, they need to be getting on these AI platforms. And probably, that is probably true in the long term from a business perspective. And at the same time, in like kind of cultural entities, I finally watched Devil Wars product too. I know I'm like very behind. AI is like kind of one of the big bads in that movie. And in a lot of like movie and TV shows that are coming out more and kind of addressing this AI revolution, AI is like the big bad and kind of the evil that they need to like overcome and push back upon. So it's just an interesting push and pull between like on the one hand, companies responding to this mass adoption or potential mass adoption at AI and so much of the culture kind of pushing back on it is this like driving force for evil. And I don't know in five years how that will look. - No, it's so true. And it's so interesting because like, I was writing about machine learning on skincare, like seven, seven years ago, something like that. - It's been around for a little long time. - Been around. - This is nothing about, I mean like this has been around. We've been, this has been in the set guys. Like we're just calling it something new nowadays. But it's like, I don't think it's quite as scary as people think, but I don't know. Maybe it is. I'm still, I'm still a late adopter. So I obviously have a little bit of apprehension. Well, I did, I did stump one of the bots. I will tell you, so I started kind of pressing in seeing like where are these guardrails? Like when will it start to say, I can't help you with that? And I was started to ask you questions about, you know, like how can I feel better? How can I mend a broken heart? How can I just kind of like really going for it? Like seeing what it would say. - What foundation? - Foundation. - Only. - I'm just broken. - The only chatbot that I stumped was the Ulta chatbot. And it's so new. - Is the world of giving you broken heart advice? - Yeah, it said the foot mask. It gave me the foot mask. It's literally. - Oh my God. - Oh my God. - It gave me a bunch of things. And the responses were actually like kind of like cute and conversational. Let me pull this up. So for example, Sephora said, you know, while beauty products can't heal a breaking heart, self-care rituals can offer you comfort. And then it recommended the foot mask, a couple of way products, a higher dose product, stuff like that. And then Amazon did something very similar, except it didn't give me any products. Amazon just said, oh, I'm so sorry to hear that. While I can't mend a broken heart, some things can make the healing process sweeter. And it started kind of going through things I should do, like watch a good movie or have some comfort food. But then when I asked Ulta, it just straight up shut me down. It said, we're here for beauty questions. Like just no. It just kind of, it's almost like I put in just like a jar bold sentence. Like it was just like no. And I'm not, I don't have a broken heart, by the way. I was just, I was just kept pushing it. But, yeah, I mean, I don't know. It's kind of a funny thing where all of these different apps are kind of stopping. Yeah, it's hilarious. Well, it's something I've been told by more AI savvy people is like the chat spots are kind of engineered in a way where like they never want to say like, I don't know. Which is also why you run the risk of so much like bad information from the chats, is they don't want to just cop to the fact that, I mean, I guess the Ulta chat, I guess, is willing to admit. Oh my God, outside of it. Keep a beauty, please. No, I'll see. Exactly. But yeah, a lot of them, they just, yeah, they rather than, rather than tell you like, I actually don't know what is the cure for a broken heart or I don't know what the best, you know, hair serum is, like they'll just, they'll tell you something. Yeah. Emily, last question here. Curious, did you learn of any tactics that brands are using to sort of show up in these results, at least now? Well, one thing again that I heard from Jason from Front Row, is that a category that does well with these kind of chat bots is your dermatological skincare. So like skincare is maybe developed by or backed by like a physician because that has this kind of air of credibility behind that and maybe more trove of like, claims behind it or terms that the chat bot can actually pull up and rely upon. And his kind of thinking was like, you know, we might see more brands pop up that have these kind of, maybe like certifications. Because again, it's this concrete thing that the chat can sort of fall back on and pull on to recommend to you. And something too that I just heard, you know, throughout the past couple of months talking to experts about this is yeah, like brands are going to need to have not just this like beautifully merchandise kind of like landing page or Instagram page with their products, but also have this wealth of like terms on their product descriptions that the chat that can then pull from. Or you know, it's, it might be very, I don't know, cluttered as well. Like again, a lot of like data and terms specifying like, it's kind of going back to the SEO thing like you were talking about Sarah of like back when, you know, as web pages we used to have to kind of type in, you know, best top 10 foundations for like oily or mature skin. Like you're using these search terms that you know, people are looking for. Now the brands have to put that into their product descriptions of like where the terms of the chat is maybe pulling from or relying upon to recommend products. Amazing. Thank you, Emily. We'll make sure to link to your original piece about all of this in the show notes and the post about this episode. I love when you join us. Thank you, everyone for listening. Thank you so much for listening to this episode of the Glossy Beauty podcast. If you enjoyed it, please subscribe and leave us a rating and a review on Apple podcasts, Spotify, wherever you get your podcast. It really does help us out. And of course, visit Glossy.co/beauty for even more covers of the beauty industry. Bye. [MUSIC PLAYING]