• 1. Overproduction is a forecasting error before it is a waste problem, which makes demand forecasting an environmental instrument as much as a commercial one.
  • 2. Data quality decides the outcome: a model inherits the quality of its inputs, and data cleaning is where analytics projects succeed or fail, not model selection.
  • 3. Better forecasts and shorter lead times reduce the volume that must be committed before demand is known, until at the limit production follows the order rather than preceding it.
  • 4. Returns are a design problem, not a logistics cost: preventing them upstream removes emissions and waste that no downstream recycling can recover.
  • 5. Materials must be compared per functional unit including the use phase, because single-parameter comparisons between fibres rank them inconsistently and mislead.
  • 6. An index is only as credible as the transparency of its method: scores produced under different methodologies are not comparable, however comparable the interface makes them look.
Last modified: Wednesday, 9 September 2026, 9:52 PM