Content Readiness and the Case for AI in Fashion and Luxury

FSW argues for the notion of content readiness for fashion and luxury brands as a bellwether for the successful use of AI content tools at scale.

When Chat GPT hit the market in 2023, few of us in fashion and luxury were prepared for the AI-everything frenzy that ensued. Many brands began launching their own public AI experiments, from Jacquemus’s AI-generated handbag cars to beauty’s adoption of VR tools and Nordstrom’s digital mirrors. Suddenly, the buzz was all about Snap’s AR filters, digital twins, avatars, AR try-on mirrors like those of Zero10, and digital product creation tools like CLO 3D, even though all of this tech has been evolving for decades.

Like all hype cycles, the AI mélée has softened in 2024. Yet, now fashion and luxury brands are faced with the real challenge of figuring out the why, what, and how of taking advantage of this generation of AI tools in a way that makes business sense and that can be readily operationalized by existing teams. Of course, tech like the DressX instant-makeover tool has a lot of potential but it is up to brands to decide where and how their innovation budgets should be spent.

Fashion and luxury brands’ usual approach to innovation seems to be to rely upon external providers to do the work and to sort out and pilot best-fit implementation. But, AI content is different. Relying upon third-party companies and teams to experiment and produce AI content for fashion and luxury brands will not work in the long run, even for brand marketing.

Why?

AI needs structure, context, and specificity within content to work well at scale. This is not to say that AI cannot be used to bring meaning to unstructured content; but rather, if you want to train AI to learn your brand voice and meaningfully contribute to content creation, your content needs to be well-organized, consistent, de-duped, and properly tagged to work well.

For most fashion and luxury brand content the issue with AI comes down to simply a matter of content readiness. Most fashion and luxury brand content is a mess, siloed across teams with little collaboration, duplicative tools, and amazingly nonexistent content operations. Think AI is going to replace all your marketers, writers, and editors? Think again. Brands should be deepening their content teams, not replacing them.

The future of AI for fashion and luxury content needs to be about starting small and using tools that make sense for the brand and its products, teams, and resources. Brands should focus on small-scale, test-and-learn pilots to figure out the best use cases for AI for their business needs and audience.

Defining “AI” for Fashion and Luxury Content

To clarify, the term “AI” is thrown around a lot without definition and is used to cover a range of tech from machine learning to generative RAG models. For this piece, we are mainly talking about the army of generative AI tools, apps, and platforms—LLMs and RAG models galore—that are purported to “master your brand voice,” “10X your content efficiency across channels,” and “streamline your customer service operations” at scale as the marketers tell us.

The future of AI content for fashion and luxury brand marketing, advertising, communications, customer service, and digital production creation is bright. However, it can be overwhelming and hard to know where to start.

Currently, there are AI content tools to help you with everything under the sun from ideating, proofreading, writing product and ad copy, taking notes during meetings, SEO/keyword generation and optimization, generating images, videos, animations, and art, structuring, planning, and organizing your social media posts, communicating and helping customers, developing websites and apps, and even auditing your website, marketing content, and customer data to assess gaps and determine opportunities.

At FSW, our team uses AI tools every day, including Grammarly for proofreading, Perplexity for research, Otter.AI for notetaking, ChatGPT for writing and cleaning R and Python code, occasionally Midjourney for images, and so on. Like most businesses, we have experimented a lot.

Assessing the right AI tools for your brand is dizzying, even for practitioners who work with content every day. One thing, however, is clear: taking a plug-and-play approach to generative AI at scale can be a costly mistake for fashion and luxury brands for whom brand image is everything.

AI Content Readiness and Why It Matters

Any time a fashion or luxury brand executive approaches a team member to ask about using AI this-or-that for a specific content or marketing task, ask this question:

Is our content ready?

Content readiness is a gauge of a brand’s operational preparedness or maturity for content technologies or strategies. Like the content maturity model we discussed in our recent white paper on the state of luxury e-commerce, an assessment of content readiness for AI is a current state evaluation of how well-prepared a brand’s content, content systems and tools, and content teams and operations are for the useful implementation of AI content tools.

Anyone can experiment with AI. That does not mean that the AI model will produce meaningful or quality outputs. Quite the contrary. Usually, it is “garbage in, garbage out.” AI learns patterns and associations from the information on which they are trained. If the data are flawed or do not represent what they think they should, then the AI will learn and repeat those flaws in its outputs. AI relies upon the patterns and signals within the content used to train it. AI is dumb in the sense that literally only does what it is told.

If your brand has content in multiple repositories or content management systems (CMSs) managed by different teams according to different or inconsistent brand and editorial standards and templates, then implementing AI tools in wider organizational contexts presents an operational content challenge. Also, if you feed an AI model or tool completely unstructured, internally inconsistent content with potential duplications, version control issues, and overlapping taxonomies, then the AI will produce lower-quality content that will not meet your brand standards.

Look, we firmly believe that automating many processes and lower-level content tasks can be a huge cost-cutter and time-saving measure for fashion and luxury brands, particularly given the impetus towards hyper-personalization for digital e-commerce experience. AI content tools have many, many use cases for fashion and luxury, such as fine-tuning loyalty programs, as Levi Strauss & Co have done, creating personalized outfit styling ideas or virtual closets like Stylitics, or making predictive product recommendations like Amazon has been doing for ages. Or even for being a virtual partner for social media copy like Lively is doing with Attentive AI.

Outside of fashion, Klarna achieved success with an AI chatbot pilot through a partnership with OpenAI. After the pilot, Klarna claims that its AI-powered chatbot’s efficiency resulted in fewer errors, a 25% decrease in repeat inquiries, and a reduction in average call time from 11 to 2 minutes. They also claim that the chatbot is “doing the job of 700 workers” and will increase profits by $40 million in 2024. What Klarna is not saying is that the brand has a robust, centralized content strategy with strong content operations and well-defined content types and templates to structure and define its repository of help content.

How to Prepare Fashion and Luxury Content for AI

Truly meaningful, scalable AI content creation begins with getting your brand’s content strategy and operations in order and cleaning up your content. AI content tools thrive upon consistent and repeatable content structures, systemic and granular guardrails, processes, and rules, and human oversight and attenuation.

Building on a four-step model nicely captured by Enterprise Knowledge, fashion and luxury brands interested in preparing their content for AI should follow four basic steps:

1. Model the knowledge or content domain.

Define or document how your content is interconnected in an ontology or system map to codify how people, tools, content, topics, and other concepts within your organization are related.

2. Clean and dedupe the content.

Content needs to be centralized into a content authoring platform and clean, with minimal content ROT (Redundant, Outdated, Trivial).

3. Add structure and standardization.

Content models and content types with reusable templates supporting your knowledge domain content map or ontology give AI consumable and semantically meaningful content.

4. Modularize the content.

Once content is structured and cleaned, break up and deconstruct the content into smaller sections based on the content model, ideally with a componentized design and stored in a content management system.

All of this said, there is no substitute for practice when it comes to new technologies like AI. Take the time to play around with different tools and assess their ease-of-use, utility, and intuitiveness. Make active use of built-in generative AI features on social media platforms like LinkedIn, Facebook, and TikTok to see how well you like them and how you may be able to apply them to specific contexts and functions within your fashion or luxury brand.