Fabric Sourcing · Causal Cost Model

Sustainable Sourcing: True Cost Analysis

Does choosing the cheapest fabric actually cost less? This model counts downstream cost (defects, delays, markdowns, carbon and returns) to find out.

Fabric sourcing causal cost model

The order log

Order data and supply chain data is collected and linked, resulting in downstream costs per unit. Every row is a purchase order. Buyers have chosen all kinds of options historically. Some of them have been labeled as the cheapest options and others as more sustainable fabric options. The more sustainable option often has ahigher unit price, but lower carbon intensity and/or lower return rate. The question this tool answers: did choosing the cheapest fabric actually cost less once all downstream costs are counted?

Order Category Complexity Volume Unit $ Alt $ Choice Defect Delay Markdown Carbon Returns Total Gap
Note: This is illustrative data. In a real deployment these numbers come from various systems: ERP, QMS for defect costs, logistics data, merchandising data, carbon accounting system, and your OMS returns module for returns waste. This is a demo of a causal inference method; the numbers shown here are placeholders.

Matched pairs

Propensity score matching finds pairs that are nearly identical on every observable confounder, so the meaningful difference is whether the buyer chose the cheapest fabric or the more sustainable option.

How "more sustainable option" is defined in this demo: higher unit price * lower carbon intensity (kg CO₂e/unit) * lower return rate. This reflects two dimensions measurable in standard ERP and emissions data today. It does not claim superiority on labour, chemicals or end-of-life without further evidence.
Pair 1 of 6 Carbon $65/tCO₂e
🔬
Key message
Raw averages can be misleading. Purchase orders differ on category, complexity and volume. Propensity-score matching isolates like-for-like pairs so that the fabric choice (cheapest vs. more sustainable) is the only meaningful difference.

Cost breakdown

Averaged across matched pairs for the selected category, each cost component is compared between the cheapest fabric and the more sustainable option. The bars show cheapest (red) vs. sustainable option (blue).

On the sustainability costs: Carbon is ERP-traceable via your emissions accounting system. Returns waste cost often sits in your OMS returns module. These data sets only require correct attribution back to the fabric decision.
💸
Key message
For some categories the unit price saving from the cheapest fabric is eroded or reversed by higher defect, delay, markdown, carbon and returns costs downstream.

The causal finding

After controlling for category, complexity and volume, here is the estimated causal effect of choosing the cheapest fabric versus the more sustainable option. The summary cards show all four categories. Click any card to see its full cost waterfall below.

Include sustainability costs:
Uncertainty:
Net causal effect by category: cheapest fabric vs. more sustainable option
⚖️
Key finding
We may find that in certain categories, the cheapest fabric costs more in total. The unit price saving is outweighed by downstream operational and sustainability costs. The case for resetting your cost threshold is clear.