Understanding Overproduction

This article was originally published in The Interline’s second Sustainability Report. To read other opinion pieces, exclusive editorials, and detailed profiles and interviews with key vendors, download the full Sustainability Report 2024 completely free of charge and ungated.

---

Overproduction is a thorny problem, and solving it requires all the pieces of the puzzle that have led us here – to a state where fashion routinely creates far more variety and volume than the consumer market or the planet can reasonably sustain – to fit together in a new configuration to find a solution. 

To better understand where overproduction originates, the scale of the problem it creates, and what it might look like to begin to reverse it, I wanted to look at three of these key pieces: policy, technology and consumers. To gain first-hand expert insights, I also spoke with Delphine Williot, Policy & Campaigns Manager at Fashion Revolution, and Ganesh Subramanian, founder of Stylumia – a company that provides AI-driven analytics tools, which could play a role in helping recalibrate demand and supply. 

Let’s start by examining the scale of the problem. It’s widely accepted that fashion is one of the biggest industries in the world – one that is a contributor to both environmental and social crises as a direct result of its scale, and one that is on a destructive trajectory with high production volumes that keep on increasing. 

What is not known, however, is exactly how many clothes the industry actually produces annually, and, of that total, how many are sold, sent to landfills, incinerated, or otherwise disposed of. This is largely due to the fact that, according to data captured by Fashion Revolution, 89% of brands do not disclose their production volume. Estimates suggest that between 80 and 150 billion garments are made every year, 8 to 60 billions of which are not sold. In Europe alone, between 254,000 and 594,000 tones of textiles are destroyed via some method each year. By even the most charitable yardstick, this is a significant amount of overproduction. Between 10% and close to 40% of all clothing is, according to these figures, effectively surplus to the market’s requirements. And as any fashion insider will tell you, a full-price sell-through rate of just 50% is now considered fairly common – and while not all of that extra stock is disposed of after being progressively discounted, at least some of it is. 

So fashion has, by and large, accepted the idea that drowning in excess products is just part of the business model. But how far will that acceptance stretch if the volume of new product introductions continues to grow at an estimated 3 times by 2050? Even before we get to the sustainability part of the argument, where will the industry draw the commercial line on making more and more?

There may, though, be a glimmer of hope that the percentage of unsold and destroyed garments might reduce even as production increases, thanks to recent policy developments such as France’s Anti-Waste Law for a Circular Economy, introduced in 2020, the EU’s ban on the destruction of returned and unsold textiles in 2023, and, more recently, the Responsible Textile Recovery Act of California. 

According to Delphine Williot, Policy & Campaigns Manager at Fashion Revolution, policies such as the EU’s ban on the destruction of unsold and returned textiles will mean that brands and retailers  “will no longer be able to get rid of their unsold goods by simply burning them”. Williot adds that these policies “will also hopefully push for better management of textile waste by ensuring that what is [currently] considered waste can find a new life”. While a step in the right direction, Williot stresses that is not clear whether this ban will lead to a decrease in overproduction by itself, since punishing the result of a problem is not necessarily the same thing as addressing its root cause. Or as Williot puts it, “the industry needs to address its overproduction problem head on, by producing less and wasting less”. 

Williot also points out the critical role of the Waste Framework Directive in addressing fashion’s overproduction issues, as it includes an Extended Producer Responsibility (ERP) scheme that goes beyond waste management. This ensures that “brands are held financially accountable for the impact of overproduction”, by paying fees “based on the quantity of products put on the market and [putting] a financial value on the potential waste they are adding to the planet”. As compelling as this legislation feels, though, Williot notes that there are still gaps to address overproduction which include lack of “concrete commitments on the setting of textile waste management and reduction targets” and “provisions to allow the transfer of EPR fees to support regions heavily impacted by used textile exports from the EU among other areas”.

Technology: the next piece of the puzzle

Overproduction consists of multiple elements: the large number of products that are produced and sold but seldom or never worn, those that are sold but later returned, overstocks that are shifted to discount racks or outlets, and obsolete products that remain unsold. The latter is primarily due to a major mismatch between supply and demand (even with planned markdowns taken into account), with forecasting done manually, at fixed intervals – ideally backed by objective data –  versus an extremely dynamic market. If fashion is going to reduce and eventually eliminate unsold inventory – absent any large-scale move to on-demand production –  improving the accuracy of demand forecasting is going to be crucial, which is where AI-driven forecasting could prove valuable. 

Consumers’ fast-evolving interests are influenced by various sources such as brands, social media, friends, and community networks, according to Ganesh Subramanian, founder of Stylumia – a company that leverages a vast influx of market data and uses AI to surface insights from it, with the goal of providing more timely guidance and customised, geography-specific trend forecasts to brands and retailers.

In addition to effectively forecasting trends and consumer interests, AI-driven forecasting could also help brands produce more accurate sizing ranges by understanding consumers’ preferences for size and fit. 

Subramanian highlighted, though, that effective AI models depend on high-quality data; hence, they “gather detailed data at the size, fit, and SKU levels, encompassing all visual and non-visual attributes, which allows us to forecast fit and size trends by geography and category”. By integrating this information with a brand’s own data, AI could provide comprehensive insights, which would enable companies to, as Subramanian puts it, “adjust their product fits based on these insights, leading to enhanced sales relevance, increased sales, and fewer returns”.

And while more sales is not a good thing, environmentally speaking, in isolation, the words “relevance” and “returns” have a major bearing on the outcome of those sales. If more products reach the market but are actually the products the market wants, and are sold at full price, not returned, and then go on to have a lasting lifespan in the circular economy, this could still be a net positive outcome despite, on the surface, being a symptom of more growth.

Using AI tools for accurate demand forecasting, then, might not only help reduce overstocks and the environmental harm caused by them, but could also benefit brands financially by saving them unnecessary costs from producing, transporting, storing, and eventually destroying unsold products. Moreover, it prevents excess stock from being sold at a discount, which is one of the most significant expenses for fashion brands and retailers. Preventing this is particularly important for luxury fashion brands to maintain their brand value. Subramanian also mentioned that through case studies, they have noted improvements “in absolute full-price sell-through rates by 5-10%, translating into an absolute  2.5% to 5% increase in profit margins which represents a 15-33% surge in profitability”.

A key question remains, though: is producing to demand the final objective? Or does the key to making the biggest difference lie in changing the shape and scope of that demand in the first place?

The power of consumers

Consumers are arguably one of the most important contributors to fashion’s cycle of overproduction. As the old saying goes: if people keep buying something, or at least expressing a signal that they might buy it, someone is going to keep making it.

Even if brands start producing to sensed demand rather than predicted demand, if those demands are excessively high, the problem still remains. Today, consumers’ demand is more heavily fuelled by social media than by traditional advertising. TikTok in particular, where trends have an alarmingly short lifespan, has become a key cog in fashion’s economic engine. As a result of this fleeting redefinition of “trend,” by the time an online order arrives, it might already be out of style, prompting consumers to place an order for the next trendy item having never – or very sparingly – worn the original. For the brand, though, this represents “success,” since each item sold to a consumer and not returned is validation that demand existed for that product. 

However, again, there is a glimmer of hope here, with a shift seemingly happening in consumers’ attitudes towards mindful consumption. In the past few years, we have seen the rise of various movements, from ‘anti-hauls’ in the early 2010s, to ‘deinfluencing’ in 2023 and ‘underconsumption core’ in 2024. One could argue that they are just like any other trend on social media with a short lifespan, but they all essentially have the same core message: conscious consumption. 

While there is undoubtedly a long way to go until these trends become more widespread and we see a real shift in consumer’ demand globally, it is key to remember that while shifts in trends happen quickly, changes in consumer culture take time. After all, we did not get here – to the place where fashion is both over-wanted and over-made – overnight, so, it is rather unrealistic to expect a quick resolution or a sudden reversal. 

But what does a potentially lower demand for products mean for fashion brands in terms of profitability? How far can growth be scaled back before it becomes inimical to business? 

As William Green of modular menswear brand LESTRANGE wrote in last year’s Sustainability Report, it means brands and retailers have to find innovative ways to work with consumers and regulators to figure out new business models. In the nearer-term it might also look like brands working to decouple their growth from production by moving away from being heavily product-focused and diversifying their offering in the form of services and experiences. These offerings will not only serve as additional revenue streams for brands, but also enable them to connect with their customers in ways not possible through products alone, and in turn to better understand the shape and scope of demand. 

While many fashion brands have started incorporating service and experience centred offerings, more is certainly needed to strike a balance between profitability and sustainability. This, coupled with new policy developments and technological advancements that enable accurate demand forecasting, can certainly contribute significantly to tackling fashion’s overproduction problem, but of course, as mentioned at the start, there are many more pieces of the puzzle that must fit together if fashion is to right-size its production to fit both the needs of people and the demands of the planet.