Abstract
Summary Publicly Available Specification 2050‐2011 (PAS 2050), the Green House Gas Product Protocol (GHGPP) standard and forthcoming guideline 14067 from the International Organization for Standardization (ISO) have helped to propel carbon footprinting from a subdiscipline of life cycle assessment (LCA) to the mainstream. However, application of carbon footprinting to large portfolios of many distinct products and services is immensely resource intensive. Even if achieved, it often fails to inform company‐wide carbon reduction strategies because footprint data are disjointed or don't cover the whole portfolio. We introduce a novel approach to generate standard‐compliant product carbon footprints (CFs) for companies with large portfolios at a fraction of previously required time and expertise. The approach was developed and validated on an LCA dataset covering 1,137 individual products from a global packaged consumer goods company. Three novel techniques work in concert in a single approach that enables practitioners to calculate thousands of footprints virtually simultaneously: (i) a uniform data structure enables footprinting all products and services by looping the same algorithm; (ii) concurrent uncertainty analysis guides practitioners to gradually improve the accuracy of only those data that materially impact the results; and (iii) a predictive model generates estimated emission factors (EFs) for materials, thereby eliminating the manual mapping of a product or service's inventory to EF databases. These autogenerated EFs enable non‐LCA experts to calculate approximate CFs and alleviate resource constraints for companies embarking on large‐scale product carbon footprinting. We discuss implementation roadmaps for companies, including further road‐testing required to evaluate the effectiveness of the approach for other product portfolios, limitations, and future improvements of the fast footprinting methodology.