Technology Is Eating Fashion
This op-ed was originally published in Business of Fashion on 31 August 2017.
Fashion companies that don’t embrace technology are sitting ducks just waiting to be picked off by sharp-shooting software companies.
Google and H&M's Ivyrevel coded couture project | Source: Courtesy
If you think you run a fashion business, you’re wrong.
A technology business with a fashion focus? Sure. Anything else and you may as well wave the white flag because the rules of the rag trade are changing. You’re either leading that change, or you’re a sitting duck ready to be picked off by a sharp-shooting tech juggernaut.
Since Amazon first started peddling books online, Jeff Bezos never once saw his company as a retailer. “Amazon is a technology company. We just happen to do retail,” said Amazon CTO Wagner Vogels in 2011. With this mentality it’s no surprise Amazon has been able to conquer retail category after retail category, solving long-old supply-chain inefficiencies using technology as the not-so-secret weapon.
From product development to distribution, nothing about the fashion supply-chain is agile. It’s impossible for traditional fashion businesses to respond to real-time demand; it takes too long to get ideas to market. Even Zara, the masters of supply-chain efficiency, can only bring a product to market in 10–15 days. In our hyper-connected digital world, a lot can change in 15 minutes let alone 15 days.
The supply-chain also fails with personalisation. Products must be designed to appeal to markets broad enough to justify producing at scale, sacrificing individualisation for unit economics. Then there’s the fit issue. Standard sizes statistically fit less than 20 per cent of the total addressable population. Too many consumers fall between the cracks of standard sizing bell-curves.
These shortcomings are being aggressively addressed by tech companies. Amazon for one has been mining its retail data and spinning up private labels to exploit product gaps discovered in the apparel market. In April 2017 the company was granted a patent for an on-demand apparel manufacturing system that creates custom clothing to the fit and specifications of individual customers. This means Amazon can not only eliminate inventory, but can respond almost instantly to market trends, and sell their products to the entire population.
Los Angeles-based Fame and Partners is another pioneer in the on-demand apparel supply chain. Like Amazon, the online womenswear label has developed a proprietary factory floor system with their manufacturing partner near Shanghai. CEO Nyree Corby says Fame and Partners use a modular design approach, allowing them to create new styles tied to their pattern and factory floor systems, which in turn maximises design flexibility, fit, and manufacturability. Corby says the rise of direct to consumer labels “translates to a larger proportion of brands now taking inventory risk than their business models previously allowed for.” She adds that reduced barriers for new fashion labels going to market “is driving fragmentation of trends and contributing to the general retail malaise.”
As consumers and their expectations digitally evolve, so too must the companies that clothe them. It’s not viable for fashion companies to design products for market segments when tech companies can design products for specific individuals. It’s not viable for fashion companies to spend weeks or months bringing products to market if tech-companies can do the same in seconds.
Technologies like data mining, machine learning, pattern bootstrapping, and product virtualisation are the tools of the new game. Tools that are already bolstering the arsenal of tech retailers like San Francisco-based Stitch Fix. They use artificial intelligence to analyse and predict purchasing behaviour and formulate new product designs based on what components of style are popular at the time. Their AI-design technology sorts through trillions of design and fabric variants to generate products that have a statistically-high chance of retail success.
Human designers cannot compete with AI-designers when it comes to synthesising complex data from multiple sources. They also can’t compete with AI-designers to action their findings and assemble, render, and launch entirely new products in seconds. A consumer may soon be browsing an eCommerce site as an AI-designer watches and learns from their actions. The machine could design, render, and display new products to the consumer in real-time based on what it believes they want. The product could then be manufactured only after the consumer has purchased the product, eliminating inventory risk.
This supply chain revolution doesn’t only apply to mass-market fashion brands. Luxury brands cannot claim superiority when tech-driven mass-market players can guarantee a more personalised and better-fitting product.
Technology also shifts the creative process towards a more symmetric interaction between consumers and brands. With AI, brands have the scalability to use individual customers as the basis of inspiration for designs. H&M’s Ivyrevel has collaborated with Google to translate “a week of your life into a one-of-a-kind design.” Lifestyle data is collected through an Ivyrevel app, including tracking venues they visit and activities they do. The app learns “who you are, what you like to do, and where you like to go,” and then proposes a unique dress design for a specific occasion.
This might sound like a novelty, however, it’s just the beginning of a movement where technology begins to inform the creative process. To remain at the cutting-edge, luxury brands must learn to harness AI to pioneer new and meaningful experiences with consumers.
Fashion businesses need to start their transition into technology companies now. The sooner they start, the sooner they’ll cultivate the domain expertise required to remain competitive in the future. Firstly, digitise historical designs and build a rich database of products split into their individual variants. When properly organised, a human or AI designer can easily reference this library to assemble a unique product without having to create anything from scratch.
Secondly, ditch standard-size grading and adopt parametric pattern grading. With parametric grading any product design can be made to fit any body type. It is getting easier and easier to capture customer body data, from taking 3D body scans on smartphones to predicting 50+ measurements from a few questions about fit. It’s only a matter of time before the mass market falls for a bespoke fit, and you don’t want to be dependent on standard sizes when that time comes.
With parametric grading and bespoke fit comes the third recommendation: supplement your mass-produced inventory with on-demand production. You can quash sizing-related problems, eliminate unsold inventory headaches, and be responsive to consumer demand on a sale-by-sale basis. A low-barrier-of-entry approach would be to leverage pre-sales as a way to collect a critical mass of orders before producing custom products at scale.
Finally, start collecting and analysing all the data that you have, such as point-of-sale data, e-commerce analytics and metrics about your customers. Whatever you have, collect it. Your biggest competitive advantage is locked away in the data that flows through your business, day in day out. Build infrastructure around your data to analyse and take action on the findings. Your business’ survival depends on it.
