Key Takeaways:
- Unlike fashion, where process and manufacturing innovation is centralised up the value chain, beauty is experimenting with the sector-specific opportunity to deploy devices and services that would have traditionally sat in clinics or in retail locations, directly to consumers.
- These experiments are receiving big business and giant enterprise backing, but the long-term outlook is difficult to predict, and proprietary at-home devices (especially ones tied into an ecosystem of consumables) in other sectors have had very polarised outcomes when they’ve clashed against existing supply chains and networks of intermediaries.
- Generative AI has hit a couple of milestones in fashion this week: new investment from Reddit royalty Alexis Ohanian in AI virtual try-on company Doji shows Valley interest in fashion technology, and Gucci’s endorsement of AI for artistic campaign purposes could be a watershed moment for a new cohort of generative AI users with speculative ambitions.
Big questions from the beauty sector about where technology resides in the product and consumer journey.
In fashion, the vast majority of direct process innovation happens far upstream from the consumer. And there’s realistically little prospect of apparel brands selling devices directly to shoppers that can make or even decorate their clothes at home.
Digital direct-to-garment printing, for instance, is capable of putting out high-quality, unit-of-one items, and printing facilities that specialise in this business model are located in-country, close to the eventual consumer, but the footprint of the hardware required means it has to live in facilities a step or two upstream. (Those facilities often package together fulfilment, distribution, and drop-shipping services, so they are not strictly single-purpose.)
And the long-term vision for all manner of digital, on-demand, and connected apparel production technology is for it to become more embedded into the existing global infrastructure of sourcing and production, where it can progressively transform product journeys for larger and larger brands.
So while fashion might be making some larger moves in wearable technology, the industry has little appetite for trying to position technology as a way for consumers to become production nodes for the things they wear.
This week, evidence continued to mount that beauty might see things differently.
Tech startup BoldHue, is a good example of this different philosophy put into practice. After a closed testing period (and after receiving $3.4 million in seed funding), the company opened up further over the last week or so, putting more of their $295 dollar starter kits into US beauty buyers’ hands, and competing for scarce space on their dressing tables and nightstands.
Designed to help consumers find the perfect shade of makeup foundation from the comfort of their own homes, the eponymous BoldHue device (a pinkish, conical box) dispenses custom-blended foundations, working from what the company calls a “universal formula,” and dispensing different amounts of something close to CMYK “inks,” to create a bespoke blend based on the customer’s facial scans.
So far, an interesting idea. And one that the giants of beauty are also placing some careful strategic bets on. Last month, L’Oréal debuted an array of technology for home use at LEAP 2025 – including its Cell BioPrint, a tabletop hardware device providing personalised skin analysis in minutes, including skin’s biological age and ingredient responsiveness (which will, we assume, naturally segue into recommendations that pull from the company’s own stable of products). The multinational also recently opened a $160 Million Research and Innovation Centre in New Jersey, although this covers both traditional innovation and what we might call this new category of “beauty innovation at the edge”.
And that “edge” really is the crux of an unspoken tension behind this sort of consumer-purchasable mechanical innovation. Fashion is no stranger to taking interesting hardware and putting it in retail or pop-up settings as a way of showcasing new processes in either production or recycling. Sneaker brands have framed in-store customisation opportunities, and snapshots of design and engineering, as reasons to visit flagships. And apparel brands have occasionally deployed recycling machinery in key locations to demonstrate their commitments.
But this is the edge in the traditional definition: at the point of retail, or in purpose-built experience centres. This is something beauty has also been pursuing: startup SmartSKN debuted a skincare lab pop-up at BSPOKE/LA, a cultural hub in Los Angeles, California, as well as in other experiential locations. SmartSKN holds the exclusive rights to introduce Korean beauty tech company Lillycover’s AI-driven skincare robots (which conduct their own workflow from AI skin analysis to personalised formulation) to the US market, but these are not consumer-ownable devices, and have closer analogues in the established network of clinics and other stakeholders in the beauty value chain.
Herein lies the tension: if more investment continues to pour into developing consumer-facing devices for home use, consumers’ interest in visiting these kinds of experience centres orintermediaries could diminish.
Already we might be seeing some initial signals from that space Cutera – the Delaware-based developer of laser and energy-based aesthetic solutions for dermatologists and medical professionals – recently filed for Chapter 11 protection in US Bankruptcy Court for the Southern District of Texas in order to restructure its debt. Is any of that directly attributable to this early push into selling skin analysis and on-demand, at-the-edge production? Unlikely. The worlds of medical devices and consumer cosmetics are still separate enough for this to be more a case of a specific company’s woes.
But are the winds potentially blowing in that direction? Perhaps. And the kind of “moments” that we see springing up around at-home solutions like microcurrent skincare tool NuFace, and the broad array of LED masks and similar self-serve treatments, certainly don’t augur well for the companies invested in keeping beauty tech enshrined in its current place upstream.
In fashion, generative AI hits two milestones in avatars and ad campaigns
Virtual try-on in fashion is nothing new, and although plenty of companies have targeted this space, few have pulled in investment from the Silicon Valley and venture capital elites.
That changed this week. Doji, a buzzy app that lets users create AI avatars in their own likeness and virtually try on clothing, caught the attention of Reddit co-founder Alexis Ohanian. Sharing his enthusiasm on X, Ohanian praised Doji for bringing a sense of fun back into the consumer market through AI innovation, and it was reported that his firm – Seven Seven Six – has made a significant investment in the company’s early-stage round.
The Interline has quietly tested Doji, and while our team found the generative AI part of the solution compelling, a lack of flexibility in body types seemed to be a limitation. Nevertheless, big venture capital doesn’t make a habit of investing in ideas that are just fun, and Doji’s backers clearly expect this to be the next wave of online shopping.
And at the same time, this week has seen something of a step-change for AI “creators” (The Interline doesn’t intend those quotation marks as snark, but as yet there isn’t a clear and workable label for this emerging cohort of people) as luxury brands began more actively endorsing AI-generated campaigns. Sybille de Saint Louvent, who describes herself as a “multidisciplinary creative designer” who previously used AI to create ‘spec’ ads for various brands (a common practice amongst agencies and artists using more traditional tools, who use these creations to land work from brands) posted a new campaign for Gucci, which she said was given permission by the brand to create, reported Business of Fashion.
This may seem like a minor story, but The Interline sees it as the start of a step-change, with AI professionals tacitly being recognised alongside traditional photographers, videographers, and 3D artists by some of the highest-profile brands in the world.
The longer-term question, of course, is whether AI is receiving this validation because of its newness and – knowing luxury – its slightly controversial edge, or whether the toolset, and the people who use it, are about to become a bigger piece of the outward-facing media machinery of fashion? (The Interline is under no illusions as to how far the roll-out of AI has progressed in a broad range of inward, behind-the-scenes use cases, as we’ll be analysing in the upcoming AI Report 2025 next quarter.)
Will this hyper-real, identifiably “AI” aesthetic lose its appeal over time? Will the next wave of trends favor imagery that feels unfiltered, organic, and intentionally imperfect? And if that happens, will AI be able to replicate even that level of authenticity?
Crucially, when the allure of purposefully, provocatively labelling a campaign as “AI” rubs off, will the detractors even be able to tell when and where it’s being used?
