image 6

Redesigning the value chain: why smart forecasting is only part of fashion’s future

The world as a whole has obviously changed a lot in the last twelve months.  But beyond COVID I genuinely feel as though the fashion industry has entered a completely new era – one that came about not just because of the short, sharp shock of a pandemic, but through the steady accumulation of market pressures, changes in consumer behaviour, and the ongoing evolution of technology.  And it’s a new era that the industry has entered almost without realising it.

Like many people around the world, I and the rest of The Interline’s team are working from home as the UK continues its third national lockdown. And for a good proportion of the world, remote working is probably going to persist long after the threat of COVID fades into the background.  This is going to have a direct effect on fashion in the form of increased sales of comfortable, casual wear, and decreased sales of formal office attire.

But the same forces keeping people at home are also adding an additional layer of change on top of the numerous transformations that the international fashion design, development, and production chain has already undergone as the fashion industry has wrestled with questions around environmental sustainability, ethical operation, and over-production and over-consumption of fast fashion.

As The Interline starts a new year of content with a focus on the future of planning and forecasting – capturing the new processes and systems that will combine to help brands and retailers decide what to bring to market – I want to make a case for why better forecasting is only part of the picture, and why I believe fashion is now ready to create to meet actual demand, rather than attempting to predict it.

And first I want to look at the problems that the past model has created.

How much fashion is lost to landfill? And why?

If you stop for a second and look at our industry from the outside, the amount of product that the fashion industry produces and then sees thrown away seems insane.  More than 90 million tonnes of clothing and textiles finds its way to landfill every year – a situation that stems from fashion’s use of non-renewable resources, high-volume production techniques that hinge on economies of scale, and consumer attitudes that have been shaped by the dominance of fast, disposable fashion.

Until recently there has been little incentive for brands and retailers to change this situation.  By stocking huge volumes of low-cost, relatively low-quality product, companies could insulate themselves from risk – adhering to the adage of “stack it high, sell it cheap!”.  The margins on these products were high, given their use of cheap synthetic materials, low cost offshore labour, and relatively simple construction.  As a former work-study engineer, who used to be responsible for costing garments by combining their bill of materials (BOM) and bill of labour (BOL), I know that target margins in the past were anywhere from 50% to 80%, depending on the complexity of the manufacturing required.  And since then, the fast fashion industry has driven those margins even higher – even if they are still eroded down to single-digit percentages by the time we have factored in the cost of goods sold (COGS), which includes marketing expenses, warehousing, logistics, mark-downs, returns etc.

Fashionable modern woman with shopping bags on landfill, consumerism versus pollution concept.

The simple fact, though, is that as the fast fashion model has risen to rule the industry, brands and retailers have become extremely ruthless about growing and maintaining those high margins.  And that has come at a huge environmental cost, especially since the only way the industry can do this is by producing huge numbers of garments on a constantly-accelerating calendar, spreading their ranges incredibly wide, and introducing new products quickly because they could be confident that shoppers were simply going to throw the old ones away and buy their latest products.

Even prior to the pandemic, though, different forces were conspiring to force a change.  Increased ethical scrutiny of supply chains was making cheap labour a false economy, while environmentally conscious consumers were beginning to rethink their willingness to buy cheap garments that came with such a high environmental pricetag attached.

But, of course, the COVID shockwave was what really threw this outdated model into the spotlight, with retailers left holding huge volumes of dead stock, and customers completely shifting their buying behaviours in weeks, not years.  And taken together, these long-running and immediate forces are, in my opinion, already combining to make the bulk-buy, low-cost model unsustainable. In more ways than one.

COVID: when order sizes crashed by 50% overnight.

Once the initial fear of COVID had taken root close to home, in the spring of 2020, shoppers and governments then began to look at the impact it had had on domestic and international supply chains.  They discovered that the textile supply chain had lost literally millions of workers, who had been laid off with little or no notice, and a global humanitarian crisis already underway.  Why?  Because consumers in Europe, the UK, and the USA had scaled back their purchasing, and retailers left with a full season’s worth of dead stock in storerooms and distribution centres had started cancelling huge numbers of contracts – going as far as to turn away and refuse to pay for orders in transit, or to mandate that suppliers take back items the retailer was unable to sell.

Eventually this situation did stabilise, when consumers began buying again through online channels, and when physical stores were allowed to reopen.  But even then, retailers were still hedging their bets and bracing for further uncertainty, and order volumes had dropped by up to 50% year over year.

Woman sew the facial medical at sewing machine. Coronavirus pandemia. Diy protective mask against virus.

In some cases, order quantities have not returned to normal, which continues to enact a devastating toll on suppliers, who have been hit hard financially, and have had to respond by reducing their headcounts.  In the future, if order sizes do not rebound, suppliers will have no choice other than increase prices – eating into the margins that brands and retailers value.  And when this happens, the industry will desperately need to find a way of smoothing out the peaks and troughs of demand, and forging a more stable relationship with talented suppliers who have been squeezed in a serious way in the last twelve months.

What’s the alternative to low cost, high volume production models?

Arriving at a demand curve that isn’t characterised by peaks and troughs is, therefore, going to require the fashion industry to move away from the high-volume, low-cost, risk-spreading model, and towards a way of working that is shaped more by measurable scientific certainty that brute force and guesswork will never be able to deliver!

But that standard of producing according to estimated demand, then discounting and disposing of excess inventory, has become deeply entrenched in the way fashion has operated for the last few decades.  Buying and merchandising processes are, as a result, almost universally managed by people manually looking back over recent years’ performance of a given style, colourway, or size ratio, and then using that past data to extrapolate likely trends for the near future.

Instead, fashion needs to be supporting buyers and merchandisers with new insights and new tools that allow them to use more relevant, even real-time data and analytics – with or without the help of artificial intelligence or machine learning.  By making this shift, buying and merchandising decisions can be modelled, simulated, and made in a matter of days or weeks based upon the proximity of your value-chain partners – rather than the weeks or months they typically take today.

But just making those decisions in a shorter time frame, and with the help of smarter tools, will only get fashion half the way to the new era.  At the same time as overhauling the way the industry plans what to create, fashion will need to re-evaluate the way it engages with its suppliers, to ensure that the speed of production can match the new speed of the market – without sacrificing sustainability and product quality.

Letting suppliers take the lead.

In the vast majority of cases, fashion retailers outsource their production to low-cost countries, and have done for decades.  As a result, a lot of manufacturing expertise has fled consumption markets, and across Asia, Africa, and other manufacturing destinations, businesses from the niche to the multinational have built their reputations on manufacturing excellence.  Obviously, there is considerable variation between different types of suppliers, but broadly speaking the apparel and footwear supplies base is far more technically accomplished than it’s typically given credit for.  And as brands and retailers have begun to realise this, they have also started to collaborate more closely with their key suppliers in a way that would best be described as co-creation, rather than the traditional method of pushing a set of technical specifications (a “tech pack”) to partners overseas, and waiting for prototypes and samples to be sent back – then iterating until the target delivery date has been agreed.

It’s worth noting that some brands and retailers have already sidestepped this model through vertical integration (wholly owning their factories) or by established very close relationships with smaller numbers of key suppliers.  And in these cases, suppliers are proving themselves to be technically advanced – using PLM and 3D as core collaboration tools – and in many cases more than capable of adapting to a shift in the model.

For suppliers that are yet to invest in those technologies, though, there needs to be a clear incentive for adoption.  Because at this moment in time, factories, as I’ve mentioned, are bearing a lot of the burden of ongoing retail uncertainty.

So what are those incentives? If fashion retail is able to move to a more data-driven demand model, manufacturers will benefit most obviously from greater stability, rather than the current “feast or famine” level of unpredictability.  And at the same time, that benefit will start to manifest itself in better quality products, fewer cut corners, more predictable time to market, reduced return rates, and more sustainable ethical and environmental practices.  It is, in effect, a “win-win” situation.

Chisinau, Moldova, March 12, 2019. Women working sewing workshop, sew children’s clothing on sewing machines. Creation of jobs, collective work

And to head off the inevitable criticism that working this way will mean higher material and labour costs, it’s already evident that consumers are willing to pay more for better-quality garments that last – provided they know that some of that additional revenue is being equitable shared with the people who made them.  Which is to say that a more predictable model, based on real data rather than estimations, is also likely to be a fairer and more equitable model as well.

But as we’ve established, the existing paradigm of offshore, out-of-sight sourcing is both entrenched and complicated.  Retailers, brands and manufacturers have very long lists of processes and dependencies, and replacing all of those with faster, more agile, more data-driven alternatives is going to require that most of them also leave the hands of human beings – who simply won’t be able to process the variety and volume of data at anything like the speed that a connected ecosystem of technologies will be required to deliver.

Which brings us to the point of action: fashion is already well on its way into a new era, but in order to survive and thrive in that era, everyone in the value chain will need to invest in new best-of-breed technologies to create a seamless, end-to-end data flow from design to disposal.

But at a time when investment is difficult for businesses that have already been struck by the pressures of the pandemic, where should you start, and how far can you go?

Getting ahead of the competition with AI & ML planning and forecasting

Some of the world’s leading retailers and brands have already begun to replace their traditional planning and forecasting tools with data analytics, artificial intelligence, and machine learning.  Supported by these tools, their buyers and designers are making smarter decisions, and forecasting based on more reliable, real-time data.

But to date the value of these tools has been reserved for the brands and retailers themselves, rather than working its way upstream.  Whereas, to cope with the new era, the same focus needs to be turned towards factories and other supplier partners, helping all players to plan more complex assortments, and respond more rapidly to fluctuations in the market.

To achieve the required level of change and agility, we will not only need to understand the constantly-changing order mix that the market demands, and how quickly it needs to be fulfilled, but also how these products can be measured in terms of minutes per garment, factory capacity, and the requirements for skilled resources, specialist machinery and so on.  Like forecasting, the world of production planning has remaining on a fairly static path for some time now, planning around roughly-predicted peaks of demand and taking account of supplier holidays and efficiency levels, when in reality the industry needs to go further – planning in partnership with suppliers to ensure that production is aligned with actual demand, rather than being forced to react at short notice.

But there’s also a step further that the industry could, and I believe should, go.

Turning the model on its head!

So far I’ve been talking about the way fashion has traditionally worked: producing products that it thinks people want, and then selling them.  The real revolution, though, will come when we reverse that flow and begin producing after the sale.

This sounds crazy until you remind yourself that it’s the way many other industries work.  From automotive to home furnishings, consumers are accustomed to placing orders for products that are only then put into production.  And while I’m not suggesting that fashion shoppers will be happy to wait 12 weeks for a garment, we can still switch to a demand-led production model by leveraging new technologies like 3D virtual twins and digital printing to bring turnaround times down to days or just a couple of weeks.

The impact this will have on margins, sustainability, supplier relationships and much more can’t be overstated.  Rather than discounting unsold garments, or seeing them end up in landfill, fashion could be producing only what it knows will sell – providing that everyone in the value chain is equipped with the right data, insight, and execution tools to make that happen.

Fashion shot. Attractive young woman with bright makeup and in stylish colorful clothes posing on a big city street.

Are we there yet?

Going fully demand-led is a step too far for many brands and retailers right now, but nothing is preventing it from being technically achievable.  The required technology is out there, and the integrations are possible to allow us to create a digitally joined-up value chain that can produce a physical output quickly when it receives a digital input in the form of a sale.

Planning and forecasting will be vital components of that picture, of course, because realistic lead times and collaborative production relationships between brand and supplier will be the only way that value chain can operate.

Now, I’m not suggesting that it’s a simple matter of just adding AI & ML, Planning & PLM with a mix of other best-of-breed technologies to solve an incredibly complex set of challenges.  And as technologically advanced as some overseas suppliers are, many of their counterparts are way behind the technology curve, meaning that the investment of time and money involved in migrating them to a model that’s fit for the new era will be significant.

So, yes, it’s far from easy and will be one of the biggest challenges ever experienced in the fashion industry for all the value-chain partners to solve.  But it’s definitely possible with today’s solution stack, and in my opinion this is the only way the industry can become more predictable, collaborative, sustainable, transparent – and of course profitable and futureproof.

Related articles