What is Sustainable Sourcing?

How can you use data analytics to select the best suppliers considering indicators for sustainability and social indicators?

Need Help?
Subscribed! Error
What is Sustainable Sourcing?

Sustainable sourcing is the process of integrating social, ethical, and environmental performance factors into supplier selection.

This involves assessing suppliers against a set of sustainability criteria, including labour rights, health and safety, environmental impact, human rights, and more.

Three candidate factories scored on the same three environmental measures with one to four stars each, water, effluent a

The aim is to minimize the negative impacts on the environment and maximize the positive social impacts.

Once you established measurable sustainability targets, you can use data analytics to assess suppliers' performances to select the best one to optimise your objective (cost reduction, CO2 emissions) while respecting social and environmental constraints.

Use data analytics to automatically select the best supplier with a mix of economic and environmental constraints

In this article, we will discover how to use data analytics to design an optimal Supply Chain Network to minimize costs and environmental impacts.

Sustainable Sourcing KPIs


Scenario: T-shirts Suppliers

Let us take the example used in the previous articles about life cycle assessment and circular economy.

You are a logistics performance manager in an international clothing group that has stores all around the world.

The circular loop drawn wide rather than square: production feeds a warehouse, blue delivery runs out to the store, oran

The company is sourcing garments, bags and accessories from factories located in Asia.

Stores are delivered from local warehouses replenished directly by factories.

How can you select t-shirt suppliers to minimize your costs while respecting environmental and social constraints?
The scope of the t-shirt life cycle assessment, drawn as a chain from raw cotton through the factory and sea freight to

Carbon emissions

The primary indicator driving your sustainability roadmap is the supply chain's carbon footprint.

This can be measured by calculating total greenhouse gas emissions, such as carbon dioxide, methane, and nitrous oxide.

💡 What should we consider?

  • Emissions during the cultivation of the cotton used for your t-shirts
  • Emissions from electricity generation and chemicals used for t-shirt production
  • Transportation of the finished goods (t-shirt) from the factory to the central distribution centre by sea-freight
The transport emissions formula on white with nothing around it, weight of goods times distance times an emission factor

With,
E_CO2: emissions in kilograms of CO2 equivalent (kgCO2eq)
W_goods: weight of the goods (Ton)
D: distance from your warehouse to the final destination(km)
F_mode: emissions factor for each transportation mode (kgCO2eq/t.km)

To learn more about transportation emissions calculation,

Supply Chain Sustainability Reporting with Python
4 steps to build an ESG reporting focusing on CO2 emissions of your Distribution Network

Energy consumption

Another indicator is the energy the supplier consumes to produce and deliver products.

It can be measured as the amount of energy used per production unit (called functional units in the Life Cycle Assessment Methodology).

💡 What should we consider?

  • Energy consumption for cotton cultivation
  • Energy consumption for spinning, weaving and dying
  • Energy consumption for transportation

Water usage

As it becomes a scarce resource, new regulations are pushing companies to redesign their processes to reduce water consumption.

With this indicator, you can constrain your sourcing and focus on suppliers that do not affect the local water supply.

💡 What should we consider?

  • Water consumption for cotton cultivation
  • Water consumption for spinning, weaving and dying

Waste generation

Measuring waste generated during the production process can help identify areas for reducing waste and adopting more sustainable production methods.

💡 What should we consider?

  • Solid waste generation during production

📊 How to organize data collection?

If you are a supplier and you need to provide this kind of reporting you can get the support of data analytics tools to automate it.

Sustainability data lake, the larger and more legible version of a33-13. Sources on the left, twelve measures inside the

A data lake can be used as a central source of unstructured and structured data (databases, Excel files, …).

You can implement automated pipelines to extract, process and analyze these data to build reports and provide insights.

For more information, have a look at this article about life cycle assessment,

What is a Life Cycle Assessment? LCA
Use Data Analytics to evaluate the environmental impacts of a fast-fashion retail product over its entire life cycle from production to disposal.

Social indicators

They include indicators for labour rights, fair pay, and safe working conditions to assess suppliers’ performance in terms of social sustainability.

You can conduct your own audits or use third-party assessments.

💡 How to Measure it?
In the scorecard below, your company assigns a weight to each of the criteria, which represents the importance of that criteria in the overall evaluation.

The social scorecard of a supplier, six weighted criteria with a score out of one hundred each and a total. Two things d

The auditors calculate a score linked to each criterion to represent the supplier’s performance.

The total score is calculated by multiplying each criterion’s score by its weight and summing the results.

Environmental indicators

They are used to assess if your suppliers are meeting environmental standards, focusing on energy efficiency, resource management, chemical usage and greenhouse gas emissions.

💡 How to Measure it?
A similar methodology can be applied using a scorecard.

The environmental scorecard of a supplier, laid out exactly like the social one in a37-05 but with six environmental cri

You can get support from governmental agencies and get checklists to support you in assessing the score for each criterion.

Sourcing Network Design

At this stage, you have collected information about your suppliers and performed audits with scoring.

You are now ready to use this data to drive decisions, design sourcing and logistics networks to deliver your stores.

Sustainable sourcing in three steps on a navy line, collect, audit, select, with what each step produces underneath: the

We will use an example with dummy data to understand how to implement a data-driven methodology to select your suppliers, considering environmental factors.

Problem Statement

You want to redefine the supply chain network for your fast-fashion retail company over the next 5 years.

A donut of monthly demand by market for the sourcing model, 4.89 million units concentrated almost entirely in two count

You have stores in 5 different markets around the world (USA, Germany, Japan, India and Brazil).

In each market, you can select a supplier to produce your t-shirts and you have the choice to source from abroad.

The two sides of the sourcing problem in one slide, demand under a navy band and supply under a black one, each with a w

You would like to select the right set of suppliers to minimise the cost of production and delivery, considering

  • The demand of each market (million units per month)
  • Fixed and variable costs of production in each country
  • Transportation costs from the factory to the market by sea freight
One market's cost built up in three coloured columns, navy production, red freight and black demand, so the reader can s

Your overseas suppliers are cheaper, but you need to pay high freight costs to deliver to your market.

The two cost inputs of the plant location model side by side. On the left the fixed monthly cost of a low and a high cap

Therefore, you need the support of linear programming to make the tradeoff for outsourcing.

Initial Solution: Minimize Costs

If you define the objective function to minimize the costs, you can easily guess that it will maximize the volume produced in India and Brazil.

Sankey diagram of production flows, three producing countries on the left fanning into five receiving countries on the r

💡 Insights

  • In this case, the increased freight costs of sourcing overseas are compensated by the low production costs of India and Brazil
Two sourcing patterns on one grey world map, drawn in green and red. Green routes run short, from a South American facto

Sustainable Supply Chain

In the context of green transformation, your company is implementing a sustainability roadmap to reduce the environmental impacts of your supply chain.

Therefore, your procurement team has contacted local suppliers in your predominant markets (USA, Japan).

The idea is to redesign the network of suppliers, considering

  • CO2 emissions from production and transportation
  • Energy and water usage for production
  • Waste generated by the factories
What would be the impact on your sourcing network?

Test 1: Energy and Water Savings Criterion

Let us experiment what would be the impact of implementing constraints on the average consumption of energy and water per unit produced.

The two environmental inputs of the same model, water and energy per unit produced by country, and they rank the countri

Your suppliers in Germany and Japan have invested in reducing the environmental impact of their operations.

You would like to add the following constraints

  • Average water consumption below 3000 L/Unit
  • Average energy usage below 685 MJ/Unit
A sankey of which country produces for which market in the solution with three active plants, India, Japan and Germany.

💡 Insights

  • The model decided to keep producing in India because of the very competitive costs
  • But it has switched from Brazil to Germany to compensate for the impact and reduce the average water consumption of water and energy
  • Germany and Japan are localizing their production

Test 2: Add CO2 emissions constraints

Transportation accounts for a large part of the CO2 footprint of your Supply Chain.

A five by five origin to destination matrix for the sourcing model, asymmetric and with a non-zero diagonal, so domestic

Therefore, you would like to reduce the impact of upstream flows by limiting the CO2 emissions of

  • Cultivation and Production per unit kg CO2e/Unit
  • Transportation by sea freight kg CO2e/Unit
  • Total Emissions must be on average below 110 kg CO2e/Unit
The same production sankey as a37-15 with Brazil opened as a fourth source, which is the visible effect of the added con

💡 Insights

  • For the first time, India has the lowest production output
  • Brazil is a solution to keep manufacturing costs low while limiting CO2 emissions
  • USA suppliers have not invested enough in sustainable production facilities to be retained for the domestic market supply

Conclusion

💡
If you have any question, feel free to ask here: Ask Your Question

You can easily add constraints in your model of supplier selection to apply your green transformation initiatives.

In our example, we can see that CO2 emissions constraints can lead to a localization (quasi-localization) of the production to reduce the impact of transportation.

Three network solutions side by side, each a world map of factory to market flows above a sankey of production by countr

As you might expect, these additional constraints affect the total production costs.

These results can trigger strategic discussions to find the right balance between sustainable sourcing and cost reduction.

Can Generative AI support this process? Yes!

To support decision-makers in applying these insights to their sustainability roadmap, we have experimented with using GenAI to enhance the product.

MCP architecture: a user request goes to Claude behind an MCP server, which calls a FastAPI service, with three scenar…
MCP Server setup with the algorithm packed in a product

The idea is to connect Claude AI to an algorithm for sustainable sourcing (packaged as a FastAPI microservice) to help us simulate scenarios.

The first real run: one sentence of instruction, a Run network tool call, and a structured report back with the objectiv
Example of simulation run

Using prompts, we teach Claude how to use the algorithm and explain the context of Supply Chain Network Optimisation.

We can then exploit its capabilities to guide us in selecting the different scenarios and analysing the results.

The two metric tables behind the dashboard, financial above and environmental below, one column per objective function w
Example of a summary of multiple scenarios

It is striking how well Claude summarises results in efficient, concise formats.

For more details, check out this complete case study and get more details in this video.

About Me

Let’s connect on LinkedIn and Twitter. I am a Supply Chain Engineer who is using data analytics to improve logistics operations and reduce costs.

If you’re looking for tailored consulting solutions to optimize your supply chain and meet sustainability goals, feel free to contact me.

Need Help?