Most conversations about AI in beauty are often from high-altitude: adoption curves, market projections, what the technology will mean by 2030. Useful, but often very abstract. At Renude, we proudly have access to something more concrete, so this quarter we dug in to give you the details.
We analyzed the customer conversations that ran through our AI Skin Advisor over the past three months, across our brand and retailer customers, and grouped them by what the shopper was actually trying to do. Five themes came out on top as the most common uses. What is striking is how neatly they map onto the funnel, and how badly a conventional site serves several of them.
1. "What should I be using?"
The largest group by some distance is the request for a personalized recommendation. It arrives in several forms: asking for a full routine, asking for a single product, or simply describing their skin and asking what to do about it.
This is the hardest question a website has ever been asked to answer, and the one it is worst equipped for. A collection page can filter and sort. It cannot listen to a description of someone's skin and reason from it. For a shopper in this group, the alternative to an advisor is usually leaving to ask someone else, which increasingly means asking a general-purpose AI that does not sell your products.
2. "I'm buying this. What goes with it?"
Second is the shopper who has settled on a product, either one already in their routine or one they have decided to try, and wants to know what to pair it with or how to build a full routine around it.
Commercially this is the most interesting of the five. The conversation starts on a specific product page, which means the intent is already established, and the shopper is explicitly asking to be sold more. This is basket building with the customer's permission, which is a rare thing. It also has to be done correctly: pairing advice that ignores an interaction between actives is worse than no advice.
3. "Is this right for my skin?"
Third is the shopper with a specific product in mind who wants to know whether it suits them. Again this typically starts on the product page, and it is squarely a middle-of-funnel question.
It is also the cheapest conversion a brand will ever get, because nothing needs to be discovered or persuaded. The product is chosen, the price is accepted, and a single unanswered doubt is the only thing between the shopper and checkout. Answer it credibly and the sale completes. Leave it and they either buy nervously or not at all.
4. "Tell me more about this product"
Fourth is top-of-funnel research, again usually starting on a product page. What is in the formulation, what will it do for my skin, how long before I see results.
Every brand believes this information is already on the page, and most of the time some version of it is. The gap is that it is written as description rather than as an answer, and a shopper with a specific question does not want to reverse-engineer it out of a paragraph of copy. This group is a useful audit of your PDPs: if people are asking an advisor what an ingredient does, the page has not said so plainly enough.
5. "What have you got with vitamin C in it?"
Fifth is ingredient-led search. Vitamin C topped the list this quarter, which is no surprise after an unusually sunny summer in the UK and the pigmentation concerns that follow it.
The seasonality here is worth noting for merchandising. Ingredient demand moves with weather, with whatever is circulating on social, and with the in-clinic treatments filtering into topical formats. Watching which ingredients people ask an advisor for is a fairly reliable early signal, because they are asking before they are buying.
What the pattern tells you
Three of the five themes start on a product detail page, which contradicts the usual assumption that an AI advisor is a homepage acquisition tool. Most of the value is sitting further down, with shoppers who have already found something and are trying to decide.
The other observation is that all five are questions a good in-store advisor answers without effort, and that a website structurally cannot. That gap is what the category has been living with since beauty moved online. In our published Nip+Fab case study, closing it produced a 16% conversion rate from open consultation to purchase, a 42% uplift in average order value, and a reduction of more than 70% in customer service tickets about products, ingredients and routines. That last figure is the one we would point at here: those tickets were the same five questions, arriving later and more expensively.
Frequently asked questions
What do people actually ask an AI skin advisor?
In our analysis of the last three months, five themes dominate: open requests for a personalized recommendation, building a routine around one chosen product, checking whether a specific product suits their skin, researching a product in more detail, and searching by ingredient.
Where on the site do these conversations start?
Three of the five most common themes begin on a product detail page rather than the homepage, which makes an advisor as much a conversion tool as a discovery one.
Which ingredients are shoppers asking about?
Vitamin C led this quarter, consistent with a sunny UK summer and the pigmentation concerns that follow. Ingredient demand is seasonal and moves quickly, which makes consultation data an early signal for merchandising.
Can consultation data improve the rest of the site?
Yes, and it is the most underused output. Recurring questions are a direct brief for your product pages: if shoppers repeatedly ask an advisor what an ingredient does or whether a product suits oily skin, the page has not answered it clearly enough.
Does an advisor only help new customers?
No. Some of the highest-value conversations come from engaged shoppers who have already chosen a product and want to build a routine around it.
See it on your own range
The full breakdown, alongside this quarter's trends and contributor pieces, is in the September 2026 Renude Insights report. You can read the September 2026 Renude Insights report here. If you would rather see the questions your own shoppers are asking, book a demo.
