Evisentra
Food

Food Footprints Are Moving From Category Averages to SKU Evidence

"Beef is high, lentils are low" was chapter one. The claims era demands chapter two: what is true of YOUR recipe, YOUR origin, YOUR cold chain.

5-min read · published 2026-07-03 · Evisentra Insights

The public understanding of food footprints was built on category averages — and honourably so. The landmark meta-analysis by Poore & Nemecek (Science, 2018), covering data from roughly 38,000 farms, made the big pattern undeniable and freely explorable through Our World in Data: what you eat matters more than where it travelled from; animal products sit higher than plant staples; the ranges within a category are enormous.

That last clause is the one the claims era turns on. The ranges within a category are enormous. An average describes the category; a printed claim describes your SKU. Those are different statements, with different evidence burdens.

Why food LCA is genuinely hard

A single SKU's footprint moves with the recipe (a few percent more dairy shifts the number), ingredient origin (the same tomato differs by country and season), farming method (practices matter more than distance), processing energy, packaging (see our packaging article — the wrap is its own claim), the cold chain (frozen logistics carry energy the average may not include), waste at every stage, and the humble serving size that quietly converts any number into a more flattering one. None of these are exotic — they are the ordinary anatomy of a food product, and each is an assumption unless evidenced.

What public data offers — and where it stops

Public databases are the right starting point, and the best of them is honest about being one. France's AGRIBALYSE (published by ADEME) offers free lifecycle data for about 2,500 finished food products — a remarkable public asset, built for the French production context, with documented methods. Used as a benchmark, it is excellent. Used as a substitute for your own recipe, origin, and process evidence on a printed claim, it quietly becomes an assumption wearing a number's clothing. Frameworks like ISO 14067 and the GHG Protocol Product Standard exist precisely to force the boundary, functional unit, and data-quality questions into the open.

The honest middle path

The future is not "every SKU gets a bespoke five-figure LCA before any claim." It is layered evidence: public benchmarks establishing the landscape, SKU-specific evidence for whatever the claim actually asserts, and a clear line between the two. A brand that can say "benchmark from public data; recipe, origin, and packaging from our own records; here is what is still assumed" is making a claim that survives questioning — which, increasingly, is the only kind worth printing.

Key takeaway
Category averages built the intuition; they cannot carry a SKU claim. The defensible pattern is layered: public benchmarks named as benchmarks, SKU-specific evidence for the claim itself, and assumptions declared.

What this means for brands

Founder’s Vision

The future Evisentra is building toward combines public benchmark data with client-specific SKU evidence — and refuses to blur them. A readiness review should show what is known, what is assumed, what is missing, and exactly which piece of evidence would most improve the claim, before a brand prints it on packaging. Public rules, public benchmarks, private evidence, clear boundaries: that is what an honest food claim is made of.

Sources & references

Check a claim Request a review Ask for a scoping call

Not legal advice. Decision-support only. Evisentra reviews claim readiness; it does not certify products or provide legal opinions. ← All Insights