Commercial aspects of sustainability

Recently I received a direct message whether there is research on the impact of sampling of clothing for sustainability and on the impact of overproduction for sustainability. My answer was that I have seen general statements on the impact of commercial practices in fashion on sustainability but very little underpinning by research. I did a little survey to find data based research, but I did not find any. With data based research I mean a representative sample of firms in order to be able to generalise to industry level. For me data based would be based on price calculations and volume forecasts and achievements. And that would be complemented by interviews. I did also answer that such a survey would require in depth cooperation of 20 firms (access to their SAP system) and a budget of 100K Euro.

Sampling and its impact

I will try to explain the matters at stake, starting from the two questions of my interlocutor. The first question is on the impact of sampling. Sampling is done in fashion at three moments. A design sample helps to visualise a product to decide whether to offer it in the collection and to produce it. It is often produced as a single item to be presented at shows and for clients in private presentations. Making such a sample is very costly because it requires small runs of fabric, and the manufacturing is labour intensive. The cost of a sample is often 4-8 times the cost of the final product. In addition it needs to be shipped by express service, adding also costs and environmental impact to the equation. With CAD techniques the process of selection options before making a sample has reduced the number of samples, but buyers often demand a visual support to make a choice. Rendering on a computer is not always a strong alternative.

I take a simple calculation model: if the cost of making a sample is 5x the cost of making a product in serie, the environmental impact is 5X times bigger than for a commercially made product. The key question is the ratio of samples against the total volume sold. Two factor counts: what is the average run size per product, and what is the ratio f samples against the number of products taken to production. This ratios tends to be more favourable for mass-market than for luxury markets, with brands in the middle. In luxury run sizes are around 200-500 items, and a higher ratio of designed products are not taken into production. In mid market an average run size is around 1000 items, and in mass-market around 10.000. In fast fashion these numbers are smaller. In luxury one sample is made for every 350 items, in mid markets 1 for every 750 and in mass markets one for every 8000. If the footprint of a sample is 5x bigger than for a commercial, the impact is between 0,08 and 1,4%. Salesman samples is another matter. The phenomenon of salesman samples is disappearing, but the impact is large since salesman samples are often made in series of 25-50 pieces. However they are rarely thrown away and sold normally or is showroom sales. Hence their environmental impact is much smaller that design samples. Production samples are often the first items in a production run, and used to verify that the production is conform to the design sample or requested modifications afterwards. This is all assumptions, and they should be verified in a thorough study. My conclusion is that the environmental impact of the design process is limited, but larger as the company size is smaller. Small is not always beautiful.

Overproduction and its impact

Then the element of overproduction. Overproduction in fashion is the result of poor forecasting or poor analysis of sales data. Overproduction is also the consequence of production and production location choices. Forecasting in fashion is very difficult as it is a fashionable and seasonal product. The sales data of one year/season does not give a precise view for the next season: the margin of error of any forecast is far above 20%. This is compounded by the fact that the majority of clothing production is made in Asia, with for many products a production lead time above 3 months. Hence it is for may retailers and brands impossible to adjust production for in season sales data. With real “fast fashion”, meaning producing based on in season data, with short lead time and nearby production, volumes can be better adjusted to expected sales. The other aspect is that fabric suppliers and clothing makers work with tapered prices: the larger the production order, the lower the price per item. The volumes differ per item, economies of scale are higher for denim than for woollen fabrics. But economies of scale do also penalise smaller brands, they have to overorder and allow overproduction more than large retailers that arrive faster at minimum scale levels. Overproduction is also related to the number of sizes per product, and having a good view on the client profile in terms of sizing is important for limiting overprodution.

The problem of overproduction is despite better ICT tools, endemic in fashion. There is no longer any structured data collection on sales at full price or at discount. But all anecdotal data (also that from my PhD thesis in 1992) suggests that the share of unsold products or products sold at discount has not declined over the last 50 years. The performance in fulfilling consumer needs is still abysmal. One third of consumer demand cannot be met (sizing/fit being a major issue), and one third of products can not be sold at forecasted prices and have to be discounted, sold in alternative channels or discarded. The advert effect of bad forecasting does not only affect the environment, but also squeezes suppliers. A historical trend over the last 50 years is that the retail margin (pre-calculated price difference between the cost of goods and the retail price) has gone from 40% to 70% in fifty years time, while the realised margin (the real gross margin made after sales) has hardly improved. Off course retailers mastering well the art of fast fashion (in terms of smaller collections, quicker changes and production loose by responding to sales data have the benefit over the classic model. Inditex against Primark. 

Towards Fair Terms of Trade

The business model of high mark-ups and high sales is for many reasons detrimental to the value chain as a whole. There is nothing new to that. Several countries had intensive studies on the inefficiency of the value chain in the first half of the 1990s. In several countries dialogue platforms were set up under the umbrella of “Effective Consumer Response”, e.g. in Germany under the name of Dialog Textil-Bekleidung. I was the initiator and executive secretary of the national clothing dialogue (1995-1999) that collapsed by lack of commitment of large retailers. To improve the efficiency of the value chain, to vastly improve the response to consumers, to enhance the fairness for weaker players and workers in the sector; a program on improving the work of the value chain is required, now at a global level. The work of IAF on Fair Terms of Trade is essential: voluntary commitments must be translated in a homogenous legal base for commercial practices. The set up of a platform on commercial practices would then justify research in current practices and how to improve them. And a more efficient value chain also creates profits to reinvest in more sustainable practices, products and production; dressing price formation and organising commercial practices is vital in the challenges ahead of us.