Four data challenges for supply chain visibility in fashion and how to tackle them

Why is data in fashion so notoriously bad? The challenges are well known and often blamed on the opaqueness and complexity of fashion’s supply chains.

To better understand why fashion supply chains are complex, let’s picture the assembly of a car versus the creation of a shirt.

The steps involved in assembling a car have remained largely unchanged over the years, with centralised production and consistent materials (with rare exceptions). When it comes to a shirt however, the sourcing of materials can be in a constant state of flux. Most garments supply chains are external to the companies that sell the finished shirt to us as consumers, meaning limited visibility into the working conditions at manufacturing sites.

In this article, we delve into the various data-obstacles faced by the industry: Handling heterogeneous data, the lack of definition in tiers and functions, scarcity of local data, and lastly, the lack of data quality standards. The issues to identify, collect, and process data are rarely discussed. But they should be a starting point to create a meaningful strategy that includes actionable steps towards fulfilling rising due diligence and reporting obligations. 

Heterogeneous data

A key challenge lies in the heterogeneous nature of data formats. Even within the same company, a lack of standardisation in data formats can make it arduous to link connections and define correlations between sets of information. Compounding this issue is a reliance on manual data entry, where local agents are left to encode information, leading towards discrepancies and inconsistencies.

Lack of definition in tiers and functions

Another challenge relates to attempting to visualise and differentiate supply chain data by supplier functions (often referred as tiers). 

The intricate functions involved in manufacturing fashion products, meaning that one supplier might be involved in different tasks that are not part of the same ‘tier’, often make it difficult to create comprehensive visual representations. This lack of clarity surrounding the functions of suppliers can impede the identification and mitigation of risks related to each particular tier. 

Scarcity of local data

A scarcity of data plagues domestic trades in fashion supply chains. While data on international trade flows is usually readily available, and even open source, it can be hard to track movements and transactions between suppliers inside the same country. 

Even when local data is available, language barriers often still exist. Documents presented in local languages with characters different from the Latin alphabet must pass a transliteration process which might alter the original intended sense, exacerbating the challenge of comprehending non-English data sources.

 Lack of data quality standards

The lack of data standards means that what can often be encountered is of low quality, with inaccuracies and outdated information due to factors such as frequent name changes among suppliers and companies. The dynamic nature of the industry makes it difficult to maintain an accurate and up-to-date database, posing obstacles for fashion businesses relying on supplier data.

How are tech companies responding to these challenges? 

It is evident that a multidisciplinary approach is crucial in understanding and revolutionising supply chain management in the fashion industry. 

For instance, one compelling solution gaining traction is the concept of ‘digital twins’. Some fashion brands are already embracing this innovative approach by creating dynamic virtual replicas of real-life environments. This enables them to predict risks and anticipate outcomes under various conditions, leading to improved monitoring capabilities.  

At Bendi, we are at the forefront of addressing these data challenges using cutting-edge machine learning along with our data expertise. We are developing methods for continuous screenings of forced labour and other Human Rights-related risks across supplier locations by drawing on contextual knowledge and analysing historical events.

Through collaborative efforts and a commitment to leveraging technology, we can shape a more resilient and responsible fashion industry, building in transparency at each step.

If you’re interested in learning more, stay tuned for our data science e-books and upcoming webinars.

Photo by Alina Grubnyak