Photo credits: Mary Kang / The Cut

Lately I have been thinking a lot about what actually influences me. Last weekend I went to a french vintage pop-up in New York. Most of the labels were very early stage and run by one to three people, with a small social media following. I discovered these brands because I follow Merci, c’est vintage on Instagram, a Parisian boutique whose aesthetic and curation of brands I appreciate, and they posted it on their story. The pop-up did not appear in my feed because I had searched for it or because anyone paid to put it there. These early-stage Paris designers reached strangers in New York through a boutique whose taste connects designer and audience.

A few days earlier, I had been listening to Snapchat CEO Evan Spiegel on ’s podcast. Spiegel spent a good portion of the interview talking about the lack of defensibility in consumer software. Facebook copied Stories, Instagram copied filters, and most of Snapchat’s other product ideas were inside competitor apps within months. The thing that cannot be replicated, Spiegel argued, is the ecosystem the platform has built around the user, including the habits, the social graph, the rituals.

The pop-up reached me through an ecosystem I am already part of, the small set of curators and shops whose eye I trust on Instagram. The same logic applies in any category that is easy to clone (ie. software, fashion, beauty, etc). A phenomenal product is the baseline, and the ecosystem the brand builds around the customer is what turns that product into a durable position. Merci’s relationship with the people who follow them and Snapchat’s relationship with its users are the same kind of asset, just at radically different scales.

A great example of this dynamic is the cohort of creator-founded brands that have launched in the past few years. The product comes second; the ecosystem comes first. Alix Earle’s Reale Actives launched in March 2026 with a teaser account that reached 400,000 followers in a week. Sami Clarke’s Form, founded in 2021, has reached $42M in cumulative revenue and 70,000 paying subscribers. Chelsea Parke Kramer’s Parke, founded in 2022, scaled to $16M in 2024 revenue without spending a dollar on paid ads, marketed entirely through Kramer’s own TikTok and Instagram. The audience these brands launched into was not something they had to acquire. It was the ecosystem they had spent years building, and the product was what the ecosystem had been waiting for them to make.

The harder question is what happens after the launch. The first drop is the easy part because the ecosystem is primed for it. The second drop has to earn its place again, and so does the third, and so does the new category two years in. Two things have to keep being true: the product has to be amazing enough to earn repeat purchase, and the founder has to keep showing up in the register the audience already trusts her in, because the audience came for a specific point of view, not a brand. Drift too far from what made her followable, and the ecosystem stops absorbing the product.

Not all founder-audiences are the same

The brands I am describing share two structural traits. The founders all built an audience before launching a product, and the brand's commercial identity is bound to a named individual to one degree or another. Beyond those two things, the brands run on very different engines.

In each case the audience came before the product. Chelsea Parke Kramer’s Parke, founded in 2022 as a reworked vintage denim brand, scaled to $16M in 2024 revenue without spending a dollar on paid ads. Jeanne Damas founded Rouje in 2016 after a decade of building a personal-style audience as a French model. Matilda Djerf founded Djerf Avenue in 2019 on the back of a 3M+ Instagram following.

The interesting move on the chart is the cluster in the bottom right. The vertical-expert creator brands (Form, Remedy, Katz-Moses) and the heritage founder-expert brands (Lauder, Arden) sit closer to each other than either does to the broad-creator quadrant, despite being separated by a century of platform history. The shared trait is real category expertise, accumulated through years inside the domain before the brand existed.

What is specific to today is the platform layer between the founder and the audience, and the financing layer behind it.

Social Media Ecosystems

What changed is how social platforms surface content. The 2010s playbook assumed brands could acquire an audience via paid social and then talk to it. That assumption no longer holds. As Jens Grede put it on Invest Like the Best in July 2025, “in a world where we have distrust generally for authority, the voice of an individual has probably never been higher.”

Grede’s argument is about where attention now flows: away from institutional brands and toward individuals the audience already trusts. Spiegel’s argument is about what gets protected when the product itself can be copied: the ecosystem the brand has built around its customer, not the feature set. The two arguments stack. Zara, H&M, Aritzia, and Quince all sell what are functionally dupes of luxury houses. The silhouette and the fabric weight are not what protect the original.

The dupe brands are also not under-distributed; some of them have very sophisticated retail and logistics. What the luxury house actually owns is an ecosystem: a heritage, a salesperson who has remembered you for ten years, a clientele relationship that turns the customer from a buyer into a participant in something larger than the product. The brand’s universe is what the customer is paying for: its references, its campaigns, the way the boutique smells, the way the bag is wrapped. The dupe cannot reproduce any of it. The dupe brand can copy the silhouette, but not the ecosystem.

Venture capital has been pricing this dynamic for the last few years. Megan Lightcap, who leads Slow Ventures’ $60M Creator Fund, calls it an inversion: “It’s community first, vertical first, and then you launch products and services on top of that, versus what we’ve historically seen, two kids with an idea, coming up with a product or company and then going to find customers for it down the road.” Imaginary Ventures takes a related approach with POV Beauty, Reale Actives, and Doting, pairing creators with veteran operators rather than betting on the creator alone. Two of the most active venture firms in the space are organized around the same underlying assumption: the ecosystem comes before the product.

The Layers of a Creator-Founded Brand

A few notes on what the layers mean in practice. Layer 3, conversion proof, is where the venture bet actually gets de-risked, because a creator who has already moved another brand’s product is a measurable distribution asset. Several beauty founders have credited Earle’s influence for putting their products on the map in the years before she launched her own. Layer 4 is operational pairing, matching creators with veteran beauty operators who own supply chain, retail, and operations while the founder owns vision and distribution. Layer 5, white-space product-market fit, is where most failed creator brands actually lose. Mikayla launched into skin prep, not foundation, despite expectations she’d enter color cosmetics. Alix launched into acne, the category most aligned with her own publicly documented skin journey. The common thread is restraint about what category the founder has actual authority in.

Layer 7 is where the durability question lives. Independence engineering means retail relationships, brand identity, and a creator network that all operate without the founder’s daily attention. The brands that have failed publicly, as Business of Fashion reported, typically failed at Layer 5 (no genuine white space, like Addison Rae’s Item Beauty) or Layer 4 (insufficient operational backbone, like Hyram Yarbro’s Selfless), and often both.

The durability question

The A-list celebrity category produces a wide range of outcomes. Skims, Rhode and Rare Beauty are scaling exceptionally well. Fenty’s North American sales are down double digits and LVMH is selling its 50% stake after Rihanna’s reduced involvement. The Honest Company has flattened revenue and gone through CEO turnover since Jessica Alba stepped back. The shared variable across the wins and the losses is the founder’s continued day-to-day investment, which is precisely the variable the current cohort cannot guarantee. The question for POV, Reale Actives, and Form is whether the operational scaffolding being built around them will be enough if and when the founder’s attention shifts.

What influences the influencers

If creators are the most efficient distribution channel for new consumer brands today, the question worth asking is who is upstream of them. Who do the creators trust, and what does that imply for the next wave?

The creator-founded brands I am most excited about share two traits. The founder has deep expertise in a specific niche, the kind of accumulated category authority Estée Lauder and Elizabeth Arden built before their brands existed. And the founder has an audience of early adopters and evangelists who already know who they are, follow closely, and have been waiting for the product to exist.

Expertise is what makes the brand survive. A board-certified dermatologist who is no longer the most-followed dermatologist on TikTok is still a dermatologist. A 25-year-old aspirational lifestyle creator at the peak of her cultural moment is operating on a clock that her audience will eventually age out of. The vertical expert can also extend naturally into adjacent products without breaking the audience’s trust. A dermatologist can credibly move from acne to body care to sun protection. A certified Pilates instructor can move from workouts to apparel to mindfulness, which is what Form has done. A master woodworker can credibly extend across the full tool category, which is what Slow Ventures’ bet on Jonathan Katz-Moses is testing.

The taste-makers that I am most curious to watch are the Substack writers. (The Love List) and (Feed Me) have built audiences whose relationship with their writing is closer to a private conversation than a follow. The expertise is sustained taste-making in long form that a reader pays for monthly, and the audience has self-selected for being willing to spend on what the writer recommends. A product line from any of them would land into an ecosystem already built for it.

The coming years will tell us which of these brands actually built ecosystems and which ones only built flashy launches. The answer will depend less on who had the largest audience at the start and more on whether the founder kept showing up in a way the audience continued to trust.

As much as I love the films celebrated at the Oscars, I equally love admiring the red carpet. Some of my favorites from this year were Renate Reinsve in Louis Vuitton, Rose Byrne in Dior, Teyana Taylor in Chanel just days after Matthieu Blazy’s Paris show, and Elle Fanning in a Givenchy by Sarah Burton gown that debuted Burton’s creative direction before the collection had even hit a runway. While appreciating the beauty of these looks, however, I was also wondering to myself, who is paying for these gowns? How does the actress decide what to wear? Are these pulled from the runway or custom made? What happens to a gown that reportedly took 750 hours to construct after its single night in the spotlight?

It turns out the red carpet is one of the most sophisticated top-of-funnel marketing operations in any industry, with a clear, traceable money trail from months-long deal negotiations to the red carpet itself. The industry measures this impact through Media Impact Value, a metric that quantifies awareness, desirability, and cultural relevance in dollar terms. The entire economy is engineered around a product that nobody buys, the custom gown, in service of the products that everybody does: fragrance, handbags, and beauty. Understanding that distinction is understanding how luxury really works.

Who Pays Whom, and How Much

The money trail starts with the ambassador deals, which involve brands paying the actresses. Louis Vuitton, Chanel, and Dior have cornered the market on dressing Hollywood’s most visible women, locking talent into long-term contracts worth millions annually. The celebrity commits to wearing the brand publicly, sitting front row at shows, starring in campaigns, and the house pays handsomely for that commitment. At the 2026 Oscars, Chanel dressed Jessie Buckley, Teyana Taylor, and Nicole Kidman, while Dior had Rose Byrne and Mikey Madison, and Louis Vuitton dressed Emma Stone, Chase Infiniti, and Wunmi Mosaku.

For talent without an ambassadorship, one-off deals can range from $100,000 to over $1 million. According to BoF, stylists earn $25,000 to $50,000 per major event plus a 20% agency fee, jewelry sponsors pay stylists $5,000 to $10,000, and makeup artists can earn up to $45,000 for crediting a single brand. In many cases, the stylist makes more per event than the celebrity. Even fragrance brands are now sponsoring hairstylists, with niche houses like Parfums de Marly paying beauty professionals to incorporate scents into red carpet “get ready with me” content.

Then there’s the very eye popping atelier investment. Chase Infiniti’s custom Louis Vuitton took over 750 hours. Elle Fanning’s Givenchy had each wisteria petal hand-sewn by Burton’s team. One stylist described three custom dresses being made for one actress for one Oscars, with two quietly shelved the day of. When you stack ambassador fees, custom production, stylist payments, and ancillary beauty deals, a major house’s total outlay on Oscars night can reach $500,000 to over $2 million.

How the Industry Measures the Return

Media Impact Value is a proprietary algorithm developed by Launchmetrics that assigns a monetary value to every post, social media interaction, and editorial story related to a brand, layering advertising value equivalents with source-based and content-based factors across over 100 attributes. The resulting dollar figure represents what the equivalent exposure would have cost as paid advertising.

The 2025 awards season generated $3.695 billion in total MIV across five broadcasts. Louis Vuitton led at $42.7 million, Chanel at $41.3 million, Dior at $35.1 million. The 2026 Oscars set a new single-event record at $225.84 million, up 30% from the prior year, with Chanel topping fashion brands at $28.5 million.

The most interesting data point is the amplification structure. Of Lisa’s Met Gala MIV for Louis Vuitton, only 6% came from her own social channels. Louis Vuitton’s owned media generated 16%. A full 78% came from what Launchmetrics calls “Indirect Echo,” media outlets, influencers, and other celebrities amplifying the moment entirely unpaid. Launchmetrics confirmed this aligns with benchmarks: approximately 77% of ambassador MIV comes from third-party amplification that costs the brand nothing. The brand pays one person to wear the dress, and the rest of the internet does the marketing for free.

What’s Actually Being Sold

The gown is the billboard, not the product. Luxury houses do not make their money on couture. They make it on fragrance, beauty, leather goods, and entry-level accessories. Dior’s Sauvage is the world’s best-selling fragrance. Chanel’s fragrance division remains one of its core revenue drivers. The red carpet creates the aspirational aura, what the industry calls the “halo effect,” that makes someone pick up a Dior lipstick or a Chanel handbag.

The conversion path is gown to Google to fragrance counter. Lyst has documented 200%+ search surges for specific designers within 24 hours of a red carpet, and those searches land on product pages for $40 lipsticks and $3,000 handbags, not couture order forms. When Lancôme signed Lupita Nyong’o as its first Black ambassador after her Oscar win, the brand was betting on visibility and desirability translating into beauty sales at scale. The custom gown costs tens of thousands to make and is never sold, but it subsidizes a media moment worth millions in equivalent advertising for the categories where luxury actually transacts.

The System Behind the Spectacle

When following the money in both directions the architecture becomes clear. It flows from the house to the celebrity, the stylist, and the atelier. Then the media value flows back from the millions of unpaid amplifiers, the TikTok creators, the fashion commentators, and editorial slideshows, who collectively generate 78% of the impact and earn nothing from the brand.

This is a system that is consolidating. The resources required to compete for a red carpet placement are prohibitive for anyone outside the conglomerate structure: the atelier hours, the fitting travel, and the very real possibility that your dress gets shelved the day of for a competitor's. The carpet is consolidating in the same way the luxury market itself is, and for the same reasons: scale advantages compound.

I will be curious to see how this evolves as social media shifts where attention concentrates. The 2026 Oscars generated record MIV despite lower TV ratings, suggesting the red carpet’s value is decoupling from the broadcast and migrating to platforms where the indirect amplifiers live. The brands have built something quite remarkable, a marketing engine where one paid placement generates millions in unpaid media, where a product nobody buys sells the products everybody does, and where 78% of the value is created by people who actively chose to participate for free.

A few months ago, I heard the Runway CEO Cristóbal Valenzuela compare AI in the creative domain to the camera in photography: the skill is in the creative direction, and the output is the idea brought to life. I've taken this seriously enough to vibe-code a monthly moodboard that pulls my screenshots, street style saves, and other visual inspiration into a custom HTML site, a running log of how my aesthetic evolves. I like gathering inspiration on my own terms rather than letting a single platform curate it for me. One of the things I love about fashion is that it's a tool for self-expression and individuality, and AI has become very useful for the curation side of that, for collecting and organizing a visual point of view. But the bridge from having a visual point of view to finding the actual products that bring it to life is still broken.

I haven’t found a platform that can take an extremely detailed prompt and generate outfit inspiration and product results that actually look good, both individually and styled together. Imagine being able to draw from the visual sensibility of stylists like Kate Young or Natasha Colvin, then surface more affordable versions of what they’d pull. AI can’t and shouldn’t replace human stylists; their access, their relationships with emerging and established designers, and their accumulated eye are what make them irreplaceable. The goal isn’t replacement but democratization, similar to what Delphi and SuperMe are doing for expertise in other domains.

Many companies are working on this, from startups like Daydream to the bigger players like Google. And yet: type “slouchy butter-soft cognac leather tote with 90s minimalism affect” into any of them, and what comes back is a grid of brown leather bags that completely misses the point. The system strips the prompt to its barest components: leather, tote, brown. Everything that made the query specific, the mood, the cultural shorthand, the visual feeling, gets discarded before the search even begins.

I’ve been curious why this is so hard to execute on, both technically and financially. What makes fashion algorithms harder to build than AI for healthcare, manufacturing, coding, or education? Fashion is a uniquely difficult domain for search, not because the compute is expensive or the models don’t exist, but because the underlying data problem is unlike almost any other industry.

How Shopping Search Works Today

Most product search still runs on keyword matching and structured attribute tagging. Product catalogs are databases with rigid schemas: brand, category, color, material, size, price. Every item gets slotted into those fields and nothing else. When someone types “90s minimalism,” the system isn’t choosing to ignore the phrase. It physically can’t store or search against that concept because there’s no field for it. Think of it as trying to file a song by how it makes you feel in a system that only has columns for genre, BPM, and key signature.

The catalog was designed to help a warehouse manage inventory and a checkout page display options, not to understand that “quiet luxury” means unlined, unhardware’d, neutral-toned, with a specific kind of slouch. Those are inferences that require cultural knowledge, and the schema was never built to hold it.

The Players

Google Lens and Google Shopping attack the problem through visual similarity. Google Lens matches what a product looks like; it’s strong on “find me this exact thing” but weak on vibe. The more interesting evolution is Gemini, Google’s multimodal model, powering what they call SGE (Search Generative Experience), an AI layer that generates shopping answers rather than returning links. For shopping, this means training on massive sets of image-text pairs so the system can start associating phrases like “quiet luxury” with specific silhouettes and palettes rather than keywords alone. The capability is real, but Google’s ad-driven business model doesn’t always incentivize serving the best organic results.

Pinterest is arguably furthest along, not because of a superior model, but because of fundamentally different training data. Their algorithm learned from years of user-curated boards, not product catalogs. When someone pins a Juergen Teller campaign photo next to a specific loafer next to a Noguchi lamp, the system learns these things are aesthetically connected despite sharing zero product attributes. Millions of people doing this over years gave Pinterest something no product database can replicate: a map of how humans actually group things by feeling.

That map is what’s called a vector space, a way of plotting items by similarity. Every image gets a set of coordinates. In normal product search, two brown leather bags sit near each other because they share attributes. In Pinterest’s version, a bag might sit near a lamp and a photograph because users consistently grouped them. “Mood” becomes a real axis, not a metaphor, which is why Pinterest returns results that feel right even when they don’t share obvious keywords.

Daydream, built by Julie Bornstein (ex-Stitch Fix, ex-Nordstrom), approaches the problem through chat-based, AI-native search. They use a multi-model architecture: separate AI models handle different parts of the process, one interpreting the natural language prompt, another analyzing visual similarity, another matching against product inventory, each optimized for its specific task. Their training data combines editorial fashion content, social media, and product catalogs, an attempt to bridge the gap between how a magazine describes a look and how a database categorizes it. With over two million products across 8,000 brands and a $50M seed round, their bet is that a vertical model fine-tuned on fashion-specific data will outperform general-purpose tools like ChatGPT or Perplexity for this use case.

Why This Is Technically Hard

Three specific bottlenecks explain why the gap between prompts and results isn’t closing.

The training data problem. Building a model that understands aesthetic concepts requires training data that encodes cultural and aesthetic knowledge: editorial content, stylist picks, mood boards, runway context. Most companies don’t have this data, and structuring it is extremely labor-intensive. Even those that do the work run into a deeper issue: subjectivity. In medical imaging, a tumor is a tumor. In fashion, whether a kitten heel reads as "dated" or "having a moment" is a matter of opinion, and that opinion shifts seasonally. The ground truth is inherently fuzzy, which means companies aren’t just paying to label data once; they’re paying to continuously recalibrate as culture moves.

The retrieval bottleneck. Even if a model perfectly understands a prompt, it can only return results from whatever product catalog it searches. If that catalog is tagged with basic attributes (color, size, material), the matching layer collapses back to keywords no matter how sophisticated the interpretation. This distinction matters: the model is the part that interprets what someone means, while the retrieval system is the part that finds matching products. The most perceptive model in the world, searching a database that only knows “brown leather tote,” can only return brown leather totes. The intelligence dies at the point of contact with the catalog.

Aesthetic understanding is relational, not absolute. A cream linen blazer reads one way over a white t-shirt and vintage Levi’s and a completely different way over a sequined top. Most shopping search treats each item as an independent object. Pinterest gets this intuitively because its data is built on how people group things together, not how products are individually categorized. For most competitors, this is the hardest bottleneck, because it requires a model that understands combinations and context, not just individual items. It’s the difference between knowing what words mean and knowing how to put them into a sentence.

Where the Money Actually Goes

The compute costs are surprisingly manageable. Companies in this space aren’t training from scratch; they’re fine-tuning existing foundation models, large pre-trained AI systems that have already learned general patterns from massive datasets. Vision-language models like CLIP and SigLIP, trained on billions of image-text pairs, already understand basic relationships between what something looks like and how it’s described. Fine-tuning means taking that general knowledge and specializing it. Mercari, the Japanese marketplace, fine-tuned SigLIP on roughly one million product listings using a single GPU to power visual similarity features. A startup fine-tuning on a few million image-text pairs could spend anywhere from tens of thousands to low six figures in compute, a real cost but not a prohibitive one compared to healthcare AI or autonomous driving, where training budgets run into the tens or hundreds of millions.

The expensive part is building the layer that actually understands aesthetics, and there are really only three paths. The first is human curation: stylists and fashion-literate annotators labeling aesthetic qualities at $40–80/hr, producing the richest data with the worst unit economics. The second is user behavior as training data, which is the Pinterest model: board curation and co-pinning patterns as a proxy for aesthetic similarity. It’s the cheapest per data point but requires millions of users already generating signals on a platform. This can’t be purchased; it has to be built. The third is synthetic labels via large language models: having GPT-4V or Gemini describe products in aesthetic language, then using those descriptions to build a search layer. A prototype dataset might run $10–50K in API costs, making it the most efficient path today, but the quality ceiling is capped by how well those models understand aesthetics, which remains limited.

Fashion AI isn’t expensive because of compute. It’s expensive because the thing being modeled, human aesthetic judgment, is subjective, culturally contingent, and constantly shifting.

Where the Moat Lives

The strategic question underneath all of this is whether the real advantage lies in the model or in the data. Almost certainly the data, which favors platforms with existing user behavior (Pinterest) over pure-play AI startups building from scratch.

That creates an uncomfortable dynamic for the startups. If ChatGPT or Google gets fashion search to 70% quality with 100x the distribution, it’s unclear whether a 95%-quality vertical tool survives as a standalone destination. Daydream and its peers are discovery layers with no inventory, no fulfillment, and no lock-in, a thin position if the upstream AI layer commoditizes.

But there’s a possibility that hasn’t fully played out. If someone can train multimodal models on curated, aesthetically rich datasets to create a search layer where “CBK inspired spring outfits” returns results not because those words appear in a product description but because the system understands what that phrase means visually, that would be something genuinely hard to replicate. The company that builds it, with the data, not just the model, will have a real moat. Whether that company is a startup, a platform, or something that doesn’t exist yet is one of the most interesting open questions in consumer AI right now.

Most major fashion houses are built around creative excellence, not technological advantage. Having evaluated B2B fashion tech companies from the enterprise software investing side, I saw this firsthand: the sales cycles were long, internal champions were hard to find, and oftentimes technology investments got treated as operational support rather than strategic capability. Historically, AI partnerships in fashion have followed that same pattern of high-profile launch, narrow scope, and quiet shelving once operational complexity sets in.

What caught my attention about PVH Corp’s OpenAI partnership wasn’t the technology itself, it was the deployment model, which suggests they understand something most fashion companies miss about why enterprise AI initiatives actually fail. This appears to be the first serious attempt to bring comprehensive AI integration to fashion operations without requiring forward-deployed engineers in every office. The thesis isn’t that PVH has better AI, it’s that they’re betting on organizational speed and cross-functional rollout as competitive advantage. If they’re right, the moat isn’t the foundation model, it’s whether you can absorb operational disruption faster than competitors can start.

The Deployment Model That Makes This Different

The press release mentions “co-creating custom AI capabilities” and “integrating OpenAI Enterprise APIs,” which sounds like standard enterprise software, but it’s not. PVH is building on three layers: OpenAI’s platform primitives (foundation models, APIs, enterprise infrastructure), industry-specific configurations (trend forecasting models, demand planning algorithms, design exploration tools tailored to fashion workflows), and PVH-specific integrations (connections to proprietary data sources, internal systems, operational processes).

This creates something that looks like a platform but operates with custom-level depth. Design teams use AI that understands Calvin Klein’s aesthetic language, planners use forecasting models trained on PVH’s historical demand patterns and current social signals, marketers deploy personalization that knows real-time inventory positions.

The deployment model sits between two extremes. Services companies like Palantir use forward-deployed engineers for deep customization, pure SaaS scales beautifully but delivers generic capabilities that don’t account for fashion’s operational complexity. PVH is attempting a hybrid where OpenAI provides platform capabilities and co-creation support while PVH’s teams do the configuration and integration work, technology transfer rather than consulting, custom outcomes built on reusable components.

This model exists because fashion has specific constraints that make both pure consulting and pure SaaS inadequate. Trend cycles move faster than traditional six-month enterprise deployments, brands need AI that understands their aesthetic language rather than generic best practices, supply chains involve dozens of countries with thousands of SKUs across seasonal inventory cycles and wholesale/retail channels where point solutions don’t connect, and in many cases, gross margins in fashion don’t support expensive Palantir-level fees.

Why Most Fashion AI Initiatives Actually Fail

Here is where the successful pilots break down at scale. To work, this requires design teams to share work-in-progress with planning so AI can analyze consumer interest signals before samples are cut, planning teams to operate with real-time data rather than quarterly forecasts built in spreadsheets, supply chain to act on smarter production signals faster than traditional lead times, and marketing to personalize based on what’s actually available rather than campaign-driven promotion schedules.

Every function has to change how they work, and the changes have to happen in coordination rather than sequentially. Designers can’t throw inspiration over the wall to planning anymore, planners can’t wait for design to finalize before starting their work, supply chain can’t operate on the traditional timeline, marketing can’t run campaigns disconnected from inventory reality.

This is the part most enterprise AI deployments underestimate. The technology works, the ROI case is clear, the pilots prove concept, but the organization doesn’t bend. Fashion companies are optimized for creative excellence and brand heritage, which means they’ve built cultures, processes, and organizational structures that protect creative autonomy and brand consistency. AI requires operational discipline, cross-functional data sharing, and process standardization, all of which conflict with how these organizations actually work and what made them successful in the first place.

PVH’s “test and learn” language suggests they understand this. You can’t force this transformation overnight, you have to earn trust function by function, use case by use case, proving value before asking teams to change entrenched workflows.

The Organizational Readiness Gap

Most fashion brands aren’t set up for this kind of deployment. Their data lives in fragmented systems, teams work in functional silos, legacy infrastructure doesn’t communicate across departments, and organizational culture treats operational changes as threats to creative autonomy rather than capability improvements. Meanwhile, OpenAI’s platform is already built, the models work, the APIs are stable, the infrastructure scales.

The constraint isn’t technical capability, it’s organizational readiness. Research shows 42% of enterprises need access to eight or more data sources just to deploy AI agents successfully, and 86% require upgrades to their existing tech stack before AI initiatives can reach production. For fashion specifically, add the complexity of seasonal product cycles, wholesale/retail channel management, and creative processes that resist standardization.

PVH’s advantage appears to be the 12-18 month head start while competitors organize their data, align their teams, and build the internal capability to actually use these tools. By the time LVMH, Kering, or Inditex get their infrastructure ready and convince their organizations to change how they work, PVH will have accumulated two years of proprietary training data on what works in their specific operational context, battle-tested workflows that teams actually use rather than resist, and compounding improvements where each season’s performance makes the next season’s models better.

The moat isn’t the technology, OpenAI will work with anyone. It’s the organizational capability and proprietary integrations built on top of it, which become harder to replicate the longer they compound. The real question is whether PVH can transform their operations faster than competitors can start their own transformations.

What to Watch

If AI-driven demand planning works, it shows up in the fundamentals. Lower markdown rates from fewer excess units, higher full-price sell-through from fewer stockouts on winning items, improved initial markup from smarter buying and production decisions. For a company operating in fashion’s promotional environment, even modest gross margin improvement would be material, especially if sustained across multiple seasons.

The other metric is inventory velocity. If AI improves forecast accuracy, inventory should grow slower than revenue or ideally shrink while revenue grows. Better turns mean lower carrying costs and higher free cash flow, the kind of operational improvement that shows up clearly in quarterly financials rather than needing explanation in earnings calls.

Longer term, if this works, PVH builds advantages that compound. Every season of AI-assisted planning generates better data about what worked and why, which makes next season’s models better while competitors starting from zero face an expanding gap. Teams that learn to work with AI develop muscle memory competitors can’t hire away. As AI touches more functions, design insights improve demand planning, demand planning improves supply chain, supply chain enables better marketing, and each integration makes the others more valuable.

The bear case is worth monitoring. Innovation theater where pilots generate press but never reach production scale, creative homogenization where AI optimizes for safety and brands lose distinctiveness, services trap where every new use case requires extensive custom development and platform advantages never materialize, organizational rejection where employees find workarounds and culture resists, or competitive catch-up where first-mover advantage proves temporary.

What This Reveals About AI Deployment

PVH’s bet reveals something about enterprise AI adoption that’s easy to miss in the technology hype. The prep work - clean data infrastructure, cross-functional alignment, change management capability - likely matters more than which foundation model or technology partner you choose. That foundation takes 6-12 months to build, and there’s no obvious way to shortcut it with budget.

PVH’s real bet appears to be on organizational speed and cross-functional rollout, not technical sophistication. I’m curious to see how their gross margins, inventory efficiency, and speed-to-market change over the next 12-24 months. If those metrics move, it validates that organizational readiness is where the real competitive advantage lives in AI adoption.

Sources:

https://www.businesswire.com/news/home/20260127435575/en/PVH-Corp.-Collaborates-with-OpenAI-to-Accelerate-Brand-Building-Journey

https://www.pvh.com/news/press-releases/pvh-corp-collaborates-with-openai-to-accelerate-brandbuilding-journey

https://us.fashionnetwork.com/news/Pvh-integrates-openai-across-calvin-klein-tommy-hilfiger,1801981.html

https://www.businessoffashion.com/articles/technology/the-state-of-fashion-2026-report-ai-automation-workforce-organisation-talent/

https://www.mckinsey.com/industries/retail/our-insights/state-of-fashion

https://modelia.ai/blog/ai-fashion-brands

https://stylitics.com/resources/blog/fashion-brands-using-ai/

a16z article “The Palantirization of Everything” by Marc Andrusko

I recently attended a lecture featuring Deborah Needleman, the former editor-in-chief of T Magazine and WSJ Style, interviewed by Spencer Bailey about her unconventional career pivot: leaving the world of Lady Gaga cover shoots and Zadie Smith profiles to learn basket weaving.

It’s not an obvious career move. But as she spoke, the beauty of the craft came into focus, revealing how real skill is built, how knowledge travels, and why craft has become more important than ever in a culture shaped by overconsumption.

Basketry is one of the only remaining crafts that cannot be made by a machine. While machines can weave flat fabric, they can’t handle the three-dimensional shaping of organic materials that vary stick by stick. Every basket you’ve ever held passed through human hands. Every piece of willow has to be soaked, sorted, graded and bent into form by someone who understands its limits with their fingertips.

Needleman pointed to an overlooked gap in everyday modern intelligence. Humans negotiate global trade, build billion-dollar companies, and access more information in their pockets than existed in the Library of Alexandria. Yet most people couldn’t explain how a toothbrush is made, how a lightbulb works, or how a radiator heats a room. We’re fluent in abstraction and almost illiterate about matter itself. She called this being “materially stupid.”

Material intelligence, knowing what willow will and won’t do, is inseparable from basketry. The craft is a dialogue with plant matter that has its own temperament. It’s also a window into a broader category of knowledge that artificial intelligence accelerates but cannot replace.

Needleman described the knowledge she’s acquiring as embodied. It lives in her hands and body, not her head. You don’t learn it by querying a model or watching a tutorial. You learn it by standing next to someone who already knows—by watching, trying, failing, and adjusting. Philosophers call this tacit knowledge: understanding that can’t be fully articulated because it isn’t primarily verbal. It transfers person to person, not processor to processor.

This concept surfaced earlier this spring at the John Singer Sargent exhibition at The Met. The show made clear that what looks effortless in Sargent’s paintings came from years of slow work: studying light, drawing constantly, learning directly from people who knew more than he did. He later told students that painting couldn’t truly be taught through explanation. Even with a model in front of him and brushes in hand, trying to describe his process was “hopeless” as abstraction. You had to watch, practice, and feel your way into the work.

We’ve grown wired for instant gratification: expecting customer service to respond instantly, getting restless when a video buffers, wanting progress at work to show up immediately. Traditional English basket makers apprenticed for seven years, repeating the same task until technique became instinct. Today, many would look at that pace and call it inefficiency.

I’m extremely eager to learn and to implement systems that reduce the cycle time to learning and impact. And AI really has collapsed the learning curve for an enormous range of complex topics; it levels the playing field in ways that would’ve seemed impossible a decade ago. Knowledge feels endlessly compressible. Until it isn’t.

Some knowledge refuses acceleration. It requires repetition, proximity to excellence, and time to reshape a person’s judgment. Sargent’s mastery didn’t come from shortcuts; it came from learning how to see and how to work, slowly. The talk reminded me of that kind of knowledge, the kind that can’t be downloaded or rushed, even in a world built to deliver everything fast.

Listening to Needleman made me realize that craft isn’t about slowness; it’s about contact. In basketry, that means years with willow and tools. In any discipline, it means staying close to the materials, the real inputs, the real signals, because that’s where judgment is formed.

Apprenticeships once forced that proximity; today we get there through tight, repeated feedback loops. The cadence changes, but the principle doesn’t.

Craft takes time not for tradition’s sake, but because materials won’t reveal themselves any faster. Willow, paint, and light all teach through sustained attention. Craft isn’t anti-speed. It’s pro-proximity. Whatever the medium, real judgment comes from repeated, attentive contact with the work itself—the kind of understanding no shortcut can produce

I’ve always been intrigued by the intermingling of content and commerce—especially in a world saturated with new brands, creators, and an endless stream of marketing. People don’t buy luxury products just because they’re made with better materials—though high quality textiles and construction should be expected. They buy them because of how those products make them feel—what they represent, the identity they help shape, the aspiration they signal.

That emotional resonance is the true currency of luxury. And it’s why I’ve become increasingly fascinated by the relationship between film and fashion—two industries that, at their best, don’t just trade in aesthetics, but in narrative and meaning.

Costume Design as Storytelling

Lately, thanks to conversations with my film-expert friend Avni (brilliant founder of Dolly Card), I’ve been thinking more about costume design as a form of storytelling. In movies and tv shows, clothing isn’t just aesthetic—it shapes perception. It tells us who a character is, how they see themselves, what they’re trying to hide or become. The right outfit can map a character’s emotional arc, reveal a shift in power, or signal cultural critique. And when those looks resonate off-screen, they create entire style movements—turning fictional characters into real-world icons.

This dynamic feels especially salient right now. As traditional media channels grow noisier and less efficient, film has emerged as a powerful medium for emotional storytelling and long-form brand expression.

Two Approaches to Film Partnership

The Production Model: YSL and Parthenope

One recent example is Parthenope, Paolo Sorrentino’s latest film, produced in collaboration with YSL Productions. The film is a sensual, surreal meditation on youth, memory, and melancholy, set in Naples. YSL’s involvement isn’t obvious—there are no logo placements or overt styling moments, which is what makes it so interesting. The house’s codes—restraint, seduction, elegance, a certain intellectual cool—are infused into the mood of the film. It feels more like emotional alignment than overt sponsorship.

This move is as strategic as it is artistic. By investing in artist-driven cinema, YSL gains cultural credibility while retaining full creative control. They bypass increasingly expensive traditional media channels, place their point of view into the cultural bloodstream, and build IP they actually own.

The Single Moment: Proenza Schouler in Materialists

While Parthenope shows how a brand can shape an entire film through creative alignment and production support, Materialists is a reminder that brands don’t need a full studio to make an impact. Sometimes, it’s a single look in a movie that cuts through.

Dakota Johnson’s cerulean blue Proenza Schouler dress struck a chord with the movie’s audience. It encapsulates the duality of Lucy, a high-powered matchmaker navigating the chaos of her personal life while maintaining control in public spaces. Against the muted palette of a formal wedding, the color and silhouette stand out by design—understated, yet striking. It reflects her poise, self-awareness, and ability to assert presence without overwhelming the room. The fact that the dress is now sold out everywhere says everything about the power of on-screen fashion to drive demand and spark desire.

The Legacy of Film-Fashion Moments

This crossover between story and style isn’t new. Breakfast at Tiffany’s made Audrey Hepburn’s Givenchy dress and pearls the blueprint for a certain kind of polished, aspirational femininity. In Anna Karenina, there’s a ballroom scene where Keira Knightley wears a dramatic black gown paired with a stunning diamond Chanel necklace—a look that’s romantic, elevated, and timeless.

This type of emotional storytelling isn’t limited to high fashion. On TikTok, I’ve seen creators restyle characters from Euphoria, Succession, even The Bear using Zara or thrifted pieces—proving that narrative styling doesn’t need to be expensive to be effective. The instinct is the same: to use clothing as a way to express identity, mood, and transformation.

The Strategic Case for Film

Fashion houses aren’t investing in this type of storytelling as cultural philanthropy. There can be a real business case. Film offers a unique opportunity to own not just product placement but emotional real estate. When brands create or back content, they gain narrative control, creative IP, and distribution leverage.

Film gives brands a rare opportunity to be strategic without feeling overly transactional. It offers a mode to articulate identity, celebrate individuality, and express what they truly stand for. It transforms clothing from something we wear into something we feel—a reflection of emotion and self-discovery.

As beauty competition grows with indie brands launching weekly, even the most innovative products must secure high-touch, defensible distribution channels that build long-term brand equity and customer loyalty. A key takeaway from the Beauty Independent Dealmaker Summit was that Sephora is now table stakes—not just for consumer discovery, but as a validation criteria for potential acquirers. Nearly every brand I’ve spoken with emphasizes how Sephora’s infrastructure, merchandising expertise, and omnichannel capabilities make it a critical retail partner.

When reflecting on Sephora’s dominance, I was reminded of a moment nine years ago when I toured Penn as a high school student. Wharton marketing professor Barbara Kahn highlighted Sephora’s experiential retail model as a key competitive advantage, and over time, the retailer has doubled down on technology, education, and data-driven merchandising to stay ahead. While Sephora remains an essential channel, brands looking to scale profitably must explore new, high-leverage distribution models that create defensibility.

I’ve been particularly interested in how brands can apply SaaS-like principles to retail—driving higher-margin, recurring revenue streams through B2B partnerships and unlocking new channels beyond Sephora, Ulta, and Amazon. In recent months, I’ve tracked several emerging strategies that forward-thinking brands are using to diversify distribution, increase LTV, and strengthen acquisition positioning.

Beauty Brands Selling B2B

Rather than relying solely on Sephora, Ulta, or DTC, some brands are successfully selling B2B to estheticians, facial chains, and national wellness franchises—establishing credibility while simultaneously securing recurring revenue streams. Sofie Pavitt’s skincare line is a prime example of this. Nearly two years ago, I discovered her now famous Mandelic Acid serum through Glowbar, a rapidly growing facial studio. If you live in a major East Coast city and are between the ages of 20 and 40, there’s a high chance you’ve heard of or seen Glowbar. Launching in a place like Glowbar provides immediate credibility from professional estheticians and a more curated retail environment, avoiding the overwhelming selection found at Sephora or Ulta. Fewer choices mean a higher likelihood of purchase, and an implied stamp of approval from experts enhances brand trust.

Another brand executing this strategy brilliantly is Sidia. The Canadian-born brand has leveraged facial chain Formula Fig as a key partner. As Sidia’s founder, Erin Kleinberg explained in a Beauty Independent interview:

"Formula Fig has been a huge brand builder for us. Their curation of brands in the retail section of the store was really special, so I reached out to owner JJ [Walsh]. We connected right away, and she put us not only on shelf but at all the sinks in the bathrooms. The mirrors all read, ‘Give a fig about your skin,’ a cheeky play on the studio’s name, so everyone was taking photos in those bathrooms and sharing them."

Beyond facialists, Sidia has been extremely strategic with placements in trendy wellness studios and boutique fitness classes. The brand has partnered with Tracy Anderson Studios, an elite fitness chain with a cult-like following. According to Erin:

"Anecdotally, people discover us in the showers at Tracy Anderson. I know there aren’t that many studios, but for some reason, everyone comes up to me and says, ‘It’s you in the shower,’ and I’m like, ‘Yeah, it’s us.’"

I think this is a brilliant play—it allows Sidia to sell in bulk on a recurring basis, while also associating itself with a highly aspirational fitness icon. Consumers in these spaces are hyper-engaged, willing to invest in premium self-care, and trust the brands their favorite wellness spaces choose to stock.

Anecdotally, I’ve heard from late-stage consumer investors that strategic acquirers are paying more attention to brands that leverage dermatologists, estheticians, and other B2B distribution channels as a new unlock for long-term growth. Given the fickle nature of younger consumers, this makes sense. As Lauren Lieberman, a managing director at Barclays, put it in a recent BoF article:

"It’s easier to say our innovation hasn’t been good enough than it is to say a consumer just keeps moving around. It’s a fickle consumer."

By securing B2B placements with trusted professionals, beauty brands can build credibility, create recurring revenue streams, and hedge against shifting DTC and retail trends—a strategy that could prove critical in the next wave of industry shakeups.

The Rise of Spa & Hotel Beauty Retail

Another emerging trend in luxury distribution is beauty brands selling through high-end hotels and spas. Westman Atelier and Augustinus Bader have executed this strategy particularly well. I first came across Westman Atelier at Rescue Spa in Rittenhouse Square, where the brand’s elegant packaging and high-quality formulations seamlessly aligned with the spa’s refined ambiance. More recently, during a wellness weekend at Canyon Ranch, Westman Atelier was the makeup of choice at the hotel spa. My mom, who typically wears minimal makeup, had a consultation with the spa’s makeup specialist, who introduced her to Westman’s entire range—she walked away with six new products!! The experiential nature of the spa setting allowed her to interact with the brand in a way that felt personal and curated, making the purchase feel intentional rather than impulsive.

Similarly, Augustinus Bader has strategically positioned itself within luxury hotels and resorts worldwide, offering exclusive treatment menus that incorporate its science-backed skincare. These partnerships create a holistic wellness experience, blending cutting-edge skincare with the elevated service culture of high-end hospitality. Beyond in-hotel offerings, Augustinus Bader has expanded into standalone boutique spas, including its Skin Labs in London and New York, further reinforcing its dominance in prestige skincare.

This distribution approach is compelling for several reasons. First, it introduces brands to affluent, experience-driven consumers in an immersive, high-touch environment, making product discovery feel organic rather than transactional. Second, by aligning with prestigious hospitality brands, beauty companies reinforce their luxury positioning while ensuring that their products are experienced in the most favorable context possible. As consumer discovery becomes more fragmented, these strategic placements offer a direct, trusted path to new audiences, solidifying long-term brand loyalty.

Technology-Driven Distribution

No conversation about new distribution channels would be complete without discussing technology’s evolving role in commerce. Substack has emerged as a prominent commerce-adjacent platform, particularly through affiliate shopping, where influencers and content creators earn a commission each time a reader makes a purchase using their link. A growing number of writers from legacy media outlets—including Amy Odell (Ex-Vogue) and Emily Sundberg (Ex-The Cut)—are leaving traditional publications to launch their own independent, paywalled newsletters, monetizing their expertise while reshaping how readers discover trends and products. This shift moves brand engagement away from mass-market advertising and toward niche, trusted voices, allowing consumers to interact with brands through personalized editorial storytelling rather than static ads.

Similarly, platforms like ShopMy are transforming influencer commerce by allowing journalists, creators, and industry insiders to curate custom storefronts featuring products they genuinely endorse. For brands, this model enables precise ROI tracking, offering insight into which creators drive the most conversions and which creative assets resonate best with consumers.

Beyond editorial-driven commerce, shoppable video is rapidly reshaping the purchase funnel. The explosive growth of TikTok Shop—where 50% of daily users now make purchases—demonstrates how algorithmic discovery, social proof, and seamless checkout can dramatically condense the traditional path to purchase. With Gen Z prioritizing product transparency, brands that thrive in this space rely on high-engagement storytelling rather than conventional advertising. Rare Beauty, for example, has seen massive success on TikTok Shop, leveraging authentic founder narratives, community-driven content, and mental health advocacy to build trust and engagement. The platform’s algorithm favors real-time, TikTok-native content, forcing brands to invest in TikTok-specific production studios rather than repurposing static content from Instagram or other channels.

Another emerging player in video commerce is Diddo, a platform that integrates AI-powered product recognition into video content, enabling users to shop what they see on screen in real time. Unlike traditional shoppable media, which relies on QR codes or second screens, Diddo’s technology offers a 100% native checkout process, allowing consumers to instantly purchase items featured in TV shows, movies, and social videos. The platform also suggests comparable, more affordable alternatives, giving consumers options while expanding brands’ reach beyond high-end luxury products.

For brands, this shift represents a new era of monetization—one where video content seamlessly integrates commerce, allowing brands to capture impulse purchases without disrupting the viewing experience. As consumer attention continues to fragment, leveraging technology to meet audiences where they are—whether in a newsletter, a viral video, or a streaming series—will be critical for the next generation of retail success.

As the beauty industry evolves, brands that rethink distribution will gain a competitive edge. Whether through B2B partnerships, luxury hospitality, or high-touch concierge services, forward-thinking brands are finding ways to create defensibility, build recurring revenue, and enhance brand equity beyond traditional retail channels. The brands that will thrive aren’t just those with great products—they’re the ones that master where and how those products are discovered, experienced, and purchased.

The Creativity Crisis in Luxury Fashion

Lately, the luxury fashion industry has been under fire for a lack of creativity. Each season feels like a remix of past hits, with brands clinging to monograms, heritage logos, and tired formulas. Consumers, journalists, and commentators alike have lamented high fashion’s absence of truly unique direction. I agree with these arguments—it feels like everywhere you turn, the same cookie-cutter aesthetics dominate. Historically, fashion has been a force for individuality and expression. So, how did we get here, and how can brands better marry creativity and commercial success?

The Commercial Pressure Trap

Commercial pressures have forced designers to prioritize global trends and sales performance over creative vision. Focusing on financial performance isn’t optional for any brand that wants to survive. However, overemphasizing safe, predictable designs can, counterintuitively, weaken brand equity and long-term financial performance.

This cycle of creative stagnation is further reinforced by how quickly trends are replicated and disseminated:

• Fast fashion brands like Zara and Shein quickly replicate high-end pieces, making them accessible to the masses

• These brands aggressively seed products and incentivize influencers to showcase “hauls,” amplifying the same trends across social media

• LLMs and trend forecasting models exacerbate this issue by generating content based on existing materials, resulting in increasingly derivative outputs

Chloe Desaulles articulates this concept well in her essay, On Loving Machines: A Case Study on Algorithms: “Culture mirrors the algorithms that determine the creative content we see online, impacting everything from streaming services to posts by small content creators. ‘Model collapse’ is a concept in machine learning that describes when AIs are recursively trained on their own output, ultimately resulting in the blending of information that flattens out distinct patterns.”

What Is Creativity?

After reading Chloe’s essay this past holiday season and feeling a bit jaded, I found myself reflecting on the core components of what creativity truly means. My favorite definition comes from the neuroscience book, The Runaway Species. The book delves into how the brain generates new ideas by transforming existing ones, emphasizing that creativity is not about creating something from nothing but about remixing, bending, and breaking familiar concepts to produce novel outcomes.

Eagleman and Brandt outline three core strategies that fuel creative thinking:

Bending: Taking an existing idea and shaping it into something new by altering its form or function.

Breaking: Deconstructing ideas into parts and recombining them in unconventional ways.

Blending: Merging different concepts to create something unique, often leading to groundbreaking innovations.

The book draws on examples from art, technology, science, and everyday life to illustrate how these processes manifest across human history—from the invention of the airplane to the evolution of music and design. The iPhone merges a notebook, telephone, map, and computer. Fenty’s creative direction fuses music, dance, makeup, and lingerie, evoking confidence, inclusivity, and beauty in novel ways.

Balancing Creativity with Financial Performance

This is the tension at the heart of every decision, from product design to marketing strategy. Brands can’t ignore trend forecasting—inventory decisions hinge on predicting what will resonate months ahead. But leaning too heavily on data can flatten originality. The key lies in creative risk-taking within a commercially aware framework.

Lessons from The Innovator’s Dilemma

Clay Christensen’s The Innovator’s Dilemma offers valuable lessons here. The theory explains why successful incumbents—established companies with dominant market positions—often fail when faced with disruptive innovations.

Incumbents typically focus on sustaining innovations: incremental improvements designed to satisfy existing, profitable customers, reinforcing their current business models. The dilemma arises because the very strategies that drive success—prioritizing sustaining innovations and established customer needs—can blind companies to disruptive threats that seem insignificant at first. Ultimately, companies can do everything “right” yet still lose their market leadership when disruptive innovations reshape the industry.

While focused on tech, its insights apply to fashion: companies must balance sustaining innovations with disruptive ones. Brands can carve out space for experimentation while maintaining their core business, like running two operating systems simultaneously—one optimized for stability, the other for exploration. The most successful brands don’t choose between creativity and commerciality; they cultivate cultures where both coexist, sometimes uncomfortably but always productively.

Two Strategies for Heritage Brands

As luxury brands face the constant tension of staying relevant without compromising their heritage, two key strategies have emerged:

1. Incubate innovation separately

Launch sub-brands or internal “startups” that allow for bold experimentation without the weight of legacy expectations:

• Gucci Vault’s avant-garde platform

• LVMH’s launch of Fenty with Rihanna to capture streetwear’s energy

• Prada’s Linea Rossa revival tapping into tech-driven athleisure

2. Embrace risk through limited capsules and collaborations

Test disruptive aesthetics and new business models without long-term commitments:

• Louis Vuitton’s iconic partnership with Supreme redefined luxury streetwear

• Moncler’s Genius Project continuously reinvents its image through monthly designer capsules

• Dior’s artist collaborations with KAWS and Daniel Arsham inject contemporary cultural relevance

These moves allow heritage brands to evolve, tapping into new audiences while preserving the exclusivity and craftsmanship that define luxury.

Personal Framework for Creative Inputs

On a more individual level, I firmly believe our creativity is shaped by our inputs—the people we surround ourselves with, the content we consume, and experiences we have. With that in mind, here are my personal tried and true ways of sparking creative energy:

Consume a diverse content diet: Spend time with people from different backgrounds, industries, and cultures.

Spend time offline: Connect with nature, read physical magazines, and go to McNally Jackson more.

Tap into your five senses: Reflect on when you feel most joyful and confident. What are you seeing, hearing, tasting, smelling, feeling?

Study history: Explore archives and the past, not to replicate but to find ideas worth reimagining for today.