Not all bestsellers are created equal. Fashion weeks have long been in charge of bringing trends to the industry. But whereas it used to be up to Creative Directors to inspire customers’ demand and steer the direction of the fashion industry as a whole, today it is individual social media influencers – like popular Tik Tokkers – who set the tone. And it’s up to brands to follow suit by maneuvering their way to customers before the next trend pops up – bouncing from one organic trend to another in a constant cycle of responsiveness.
Is this the way it has to be?
Producing what is trending is the essence of the fashion business. The way that our human brains are wired is what makes us want what’s new, and then reach into our wallets to get it. But, due to its contribution to the industry’s slice of the annual global carbon budget pie, this model of chasing emerging trends at the expense of sustainability is undergoing new scrutiny – especially in light of the industry’s target to cut emissions in half by 2030. Nearly 10% of global CO2 emissions come from the fashion industry, caused by a reliance on fossil fuels amongst others, from which the majority of (fast) fashion textiles are derived (polyester, nylon, acrylic, elastane, and the likes). And as trends emerge faster, in greater variety, that impact is set to increase.
Nobody is suggesting that fashion should abandon the idea of trend entirely, but at the same time the industry faces a difficult question: how does it respond to demand without sacrificing its responsibility to the environment?
The growing demand from activists for (fast) fashion brands to be held accountable for their environmentally-harmful, exploitative supply chains will have big implications for the behemoths of our industry. In the fashion capital of New York City, a new act is in development to put brands of this caliber in their place and make them pay for the damages. The Fashion Sustainability and Social Accountability Act (or Fashion Act) will hold brands with a minimum annual revenue of $100 million accountable for malpractices that disadvantage the environment and that are modelling on unethical treatment of those who make our clothes. And that act is likely to be just the start: responsibility is not something that fashion can ignore.
The Origin of Waste
Not all fashion that is produced is consumed. The leftovers eventually end up either in incinerators or in landfills, contributing to global warming. Furthermore, Oxfam reports that in the UK alone about 13 million items of clothing go to landfills every week. One reason behind this is influencers wearing an outfit only once to be posted on Instagram and then discarding it – with brands then also following suit and creating products inspired by that rapidly-fading look. This is a problem that countries in the Global South are experiencing harm from in an acute way, but it’s also a global trend. What happens next is a correlation between the transience of trend, the amount that fast fashion brands produce, and the amount of textile waste ending up in landfills. Referring back to the need for systemic change to save the planet, fashion needs to design out waste from its supply chains and take responsibility for all of its involved stakeholders, but under the current model this seems like a significant hill for the industry to climb.
Fast fashion, as a business model, relies on mass production, low prices and large volumes of sales. Democratising fashion has been the mission of fast fashion brands like the behemoths H&M and INDITEX (ZARA), and this vision hinges on those companies’ ability to produce a lot of new styles very quickly. But this degree of mass production for mass consumption has led to faster decay of the environment (biodiversity loss, soil degradation, air pollution, water contamination) as well as exploitation of human beings (garment workers – see the 2015 documentary The True Cost). Despite publicly raised concerns, including the idea that Gen Z is “killing fast fashion”, the exploitation of the planet and people to benefit a few billionaires, appears to be gaining ground rather than losing it. The juggernaut SHEIN publishes 6,000 styles per day and does not seem set to slow down any time soon, if rumours of an imminent IPO are to be believed.
Despite its detrimental impact, this real-time fashion model has had one positive outcome – it’s put data-driven design processes front and center. Fast fashion leaders like SHEIN excel in agile manufacturing, with the help of trend predicting AI algorithms that are fed by data scraped from TikTok and other services. A few years ago it was The Economist that stated: Data is The New Oil. Whereas oil used to be the fuel for the economy, particularly for fashion brands who belong to the Synthetics Anonymous group, data is now the most valuable resource – and fast fashion companies are sitting on significant reserves.
Through the lens of responsibility, though, overproduction is also a massive threat to businesses’ profitability. Consumers, regulators, and investors are becoming increasingly aware of the role that a model that’s wasteful by design is playing in fashion’s contribution to global pollution. But if the industry can extract the positive elements of fast fashion – its reliance on data – without carrying forward its insistence on physically producing every possibility, overproduction could actually be contained and smarter supply would become realistic, directly reducing levels of unsold stock.
A New Trend
While fashion has always pushed innovation forward in terms of designs, our industry has been notoriously slow in embracing technology. This is ironic given that the story of technology is, in fact, the story of textiles and industrial revolution. The Pandemic, however, has evidently acted as a catalyst for the adoption of new technologies, like virtual prototyping, robots that cut and sew fabric, 3D knitting, AR filters, NFTs and AI algorithms that predict style trends. But in spite of that gold rush towards new (and established) technologies, data-driven design strategies have yet to benefit from the pandemic boost. But if the commercial argument for becoming more data-driven wasn’t enough, then the environmental one will soon be: With the globe heating up and desperate times for humanity (or Code RED as the IPCC concludes), the time for change is well and truly upon us.
Against that backdrop the adoption of technology is no longer a luxury option, but the new normal – especially for the pollutive fast fashion sector.
But how can fast fashion temper its current model of constant, high-volume production without losing its ability to create on-trend products that its customers want to buy? Can they – and other brands – take their lakes of customer data, make use of AI algorithms that predict style trends, and let machine learning do what a design team does?
When Alan Turing cracked the Enigma code, the foundation for artificial intelligence was laid. The Turing Test, originally called the Imitation Game, serves to test if a machine can exhibit intelligent behavior equivalent to or indistinguishable from that of a human. In so many areas of personal and professional lives, that test has been passed . We are already living in the Age of AI. It is only a matter of time before the majority of manual labour is replaced by computers. And creative processes such as design may be next. In fact, many experts believe that greater reliance on AI in both manual and creative tasks could be necessary to adopt more sustainable supply chains.
So how can brands respond to customer demand based on small trend windows in a way that is both environmental and socially responsible? AI could hold the answer, so I spoke to two experts in the field of AI and fashion design to understand how two of the biggest industry-level trends in fashion – machine learning and sustainability – are intertwined.
Contacting Portugal-based DeepTech startup Fashable, I interviewed CEO Orlando Ribas Fernandes, and Chief of Fashion Susana Marques – the latter of whom is also a PhD candidate in FashionTech. Fashable uses AI to support fashion brands in creating original content and reducing waste from their supply chains. Speaking of the Imitation Game, an example of their work is visible in the image below, where one dress is created by a human and the other three by AI. [Those of you curious to find out which is the human design, scroll down to the bottom of this article.]
Whilst the rise of Amazon’s AI fashion designer has given some human fashion designers nightmares, Fashable has a different definition of what it means to design with AI. As Orlando puts it: AI is giving fashion designers superpowers, as they are served with creations they would not have come up with themselves, within a short amount of time, and asked to apply their creative judgment based on those informed suggestions. The Fashable tool generates 10,000 original designs within 10 minutes. But despite that high volume, their solution never copies a concept, so truly new designs are made visible in the blink of an eye.
Susana, a fashion designer herself, sees the potential of this approach as being “incredible and scary”. The incredible part referring to the inventiveness that the AI enables, and the ways in which it can empower the designer, and the scariness referring to the speed at which Fashable generates a large pool of original content, which will understandably give some designers pause.
Crucially, as well as saving time, the AI is not blocked by biases or limited to the fashion designer’s comfort zone therewith exploring new aesthetics.
Apart from the creative benefit that AI offers, there are other advantages to adopting this technology. With trend windows becoming smaller and smaller due to the rate at which they are conceived, near-time intelligence is becoming vital for brands’ supply chain management strategies.
With faster creation of garments, that despite being “fake” look just as real, brands can test demand from their customer base much quicker. In the simple case where customer demand is low, a brand has only lost time – not the capital ordinarily spent on producing and distributing products to retail in physical stores and warehouses. With on-demand manufacturing evangelised as a solution to pollution, this type of demand testing may be the way forward we’re looking for as an industry – a method to target micro-trends without structuring our entire production apparatus around them, with all the environmental and ethical damage that entails.
Thanks to AI helping brands to save time in the fashion cycle, it is interesting to note that another benefit of a faster and cheaper design process is the budget left for manufacturing. AI design, when combined with on-demand manufacturing, can make more capital available before production starts. Rather than having to offshore this task to low-wage countries far away from home, the opportunity to produce more locally and ethically does no longer seem unattainable.
If we have learnt one lesson from the 2013 Rana Plaza catastrophe it is that offshoring production and losing control over your supply chain should be a last resort commercially, and an inconceivable option, ethically speaking. Responding to trends and maintaining high margins has been the feedstock for subcontracting in the supply chain, causing opacity to the brand and its customers. Susana and Orlando further explain that, by saving costs on the design and marketing process, brands end up financially stronger and nearshoring becomes a smart business strategy.
Imagine that your brand is based in Europe and previously outsourced production to a country in Asia. The CO2 emissions produced from transportation alone are staggering and could easily be reduced by producing closer to home, and closer to your customers. With faster product-market fit, you will not only save costs by no longer manufacturing what doesn’t sell, you will now have space to elevate your supply chain to an ethical one, on the same continent.
In this article we have explored how our industry might be able to respond to demand without sacrificing its responsibility to the environment. Despite responding to trends being a phenomenon that will continue, the traditional methods that lead to filled up landfills are already living on borrowed time. The time for disruption has arrived and the sooner we embrace the solutions that already exist, the faster we will discontinue the pollution. We need to be open to technology, assess it carefully and adopt those solutions that will bring the industry as a whole – and your brand – closer to achieving the Science Based Targets.
Producing more just to hit MOQs and increase the chances of having bestsellers is not only incompatible with what the world demands from fashion – it’s also potentially unnecessary. Solutions to automate the design process may receive resistance at first, but as evidenced by early-adopting fashion designers, from a creative perspective it can rather enrich the experience, push them beyond their boundaries, and lead to the creation of bestsellers faster than the traditional approach. In the words of my idol Dr. Kate Raworth :
“We need to meet the needs of all within the means of the planet”.
The need for trends will remain, but the way we respond to them will be nothing like business as usual.
Answer to the question which dress is not created by AI: number 3.