Drive Sustainability at Scale

A PCF and a secondary emissions factor are not directly comparable. Different boundaries and allocation methods mean the same product can legitimately produce different numbers, and confusing the two leads to false confidence in the data.

Supplier PCFs are the only way to show real decarbonisation. No database will ever have an emissions factor for chocolate made from renewable electricity at a specific factory; only a supplier-updated PCF captures that kind of change.

Hybrid splicing is a practical middle ground. Zeroing out one component of a secondary emissions factor when you have real supplier data for it can deliver meaningful specificity without requiring a full PCF from every supplier.

Carbon is a proxy, not the full picture. Switching to a lower-carbon supplier that operates in a water-scarce region can make the overall environmental impact worse; beyond-carbon reporting is coming, even if the infrastructure is not ready yet.

‍

Emerging market consumers are more concerned about sustainability than developed market ones. The reason is proximity: floods, droughts, and crop failures are immediate realities, not abstract global anxieties.

Sustainable behaviors are rising, but consumers do not call them that. Purchases of organic and natural products are growing fast; consumers frame those choices as health and quality decisions, not environmental ones.

The green premium is not a reliable pricing lever. Direct monetisation of sustainability is weak at scale; the real opportunity is indirect, through guilt-free impulse purchase dynamics.

Sustainability branding can actively hurt sales. When environmental credentials are the headline, some consumers assume the product underperforms on its core function, the paper straw effect.

Identity matching is the next frontier, accelerated by AI discovery. Brands that align their product ideology with how a specific consumer sees themselves will unlock loyalty, duration, and price tolerance that credentials-first marketing cannot deliver.

‍

PCFs bridge two worlds that never talked to each other. Corporate footprinting and life cycle assessment developed in parallel for decades; the PCF is finally making them reconcilable at the product level.

The main practical driver is Scope 3 specificity. Companies are not adopting PCFs for the theory; they want to replace generic sector-average emissions factors with supplier-specific data that makes their footprints more defensible.

False precision is still the fundamental problem. Three providers will give you three different PCFs for the same product; internal consistency against your own baseline matters more than cross-organisation comparability.

For suppliers, timing is the strategic variable. The cost of PCF compliance is coming regardless; the difference between trailblazers, fast followers, and late compliers is whether they capture the commercial upside or just absorb the burden.

‍

Water is not fungible like carbon. Where you withdraw or pollute water matters enormously; aggregating water impacts into a single number obscures the very things that make it a real business risk.

Permitting is the existential water risk, not disclosure. Omnibus has softened reporting requirements, but water permits are tightening; a company that loses a permit cannot manufacture, full stop.

Efficiency metrics can be misleading. Improving water use per unit of crop often just increases output from the same abstraction; if basin-level scarcity is the problem, you need to track basin-level withdrawal, not efficiency ratios.

Manage water through a risk lens first. Focusing on permitting exposure, reputational risk in water-stressed regions, and supply chain basin mapping will naturally drive better water outcomes without needing disclosure requirements to justify the work.

‍

Organic controls inputs; regen measures outcomes; rewilding surrenders the land entirely. They are related but distinct, and conflating them leads to muddled strategy and unrealistic expectations.

Regen is more scalable than organic for large-scale farming. Its principles are already familiar to farmers, its standards are less prescriptive, and the productivity case for improving soil health is increasingly self-evident as degradation compounds.

Rewilding works best where agriculture is economically marginal. The UK can afford to subsidise it; the US, Canada, and Australia cannot make the same trade-off without significant economic consequences.

Premiumisation is the other viable path for rewilding, but the category has to fit. Bread probably never gets there; cocoa, as it becomes an increasingly scarce and premium commodity, might actually be a genuine candidate.

‍

The brand battle was unwinnable from the start. Meat and dairy carry millennia of cultural positioning that cannot be disrupted in a two-year marketing cycle, regardless of how good the product engineering becomes.

The product category was aimed at the wrong customer. Existing non-meat eaters do not want something that tastes like meat, and confirmed meat eaters are too deeply attached to the original to switch easily.

B2B is the more natural home for this technology. Selling plant-based protein as an ingredient into supplements, snacks, and animal feed sidesteps the consumer branding problem entirely and plays to the engineering strengths of these companies.

Emerging markets are the real growth story. High existing rates of plant-based eating across South and Southeast Asia make these markets far more receptive than North America, without requiring any conversion narrative at all.

‍

Cocoa supply faces a structural, not cyclical, crisis. Aging trees, incurable disease, artisanal mining, and agroclimatic zone shift are all compounding simultaneously with no clear systemic fix in sight.

Farmers capture under 7% of cocoa's value yet absorb most of its risk. Administered pricing in Côte d'Ivoire and Ghana means they miss price peaks and get caught at troughs, making economic resilience almost impossible.

Demand is adjusting structurally, not temporarily. Reformulation, shrinkflation, and substitution of cocoa butter are likely permanent shifts, removing the commercial signal that could otherwise incentivise supply investment.

Real cocoa is becoming a premium commodity. The most plausible equilibrium is a bifurcated market where mass-market products move to alternatives and genuine cocoa commands a significant premium, at much lower volumes.

‍

Agentic AI makes sustainability work automatable at scale. Predictable, cyclical, subject-matter-dependent workflows are exactly what agents are built for, and sustainability teams fit that description almost perfectly.

‍

The new benchmark is 80% value at 60% of the cost. Organisations will choose internal agentification or AI native service providers; either way, full-team sustainability functions will struggle to justify their cost.

‍

Deep subject matter expertise is being commoditised. The person who knew the arcane rules cold is being replaced by a well-prompted LLM; what remains valuable is systems thinking, stakeholder judgment, and knowing how to design and maintain the agent infrastructure.

‍

The manager-analyst layer is under structural pressure. Agents reduce the need for information gatekeeping and decision translation, which are the primary reasons those layers exist; managers should get ahead of this, analysts should learn to demonstrate they don't need the layer above them.

‍

Don't be the translator nobody remembers. Sustainability's value has partly been about bridging between worlds; as AI absorbs that bridging work, professionals need to identify what only they can do.

‍

SBTi V2 is a deliberate market expansion play: Category A vs B categorisation, wider target optionality, and emerging market accommodation are all designed to drive adoption well beyond the original large-multinational base.

SBTi is now explicitly pro-climate finance: For the first time, the standard takes a clear ideological position: large companies should own their residual emissions financially, with mandatory offset commitments starting at 1% in 2035 and scaling to 100% by net zero year.

More optionality means less comparability: Two companies in the same sector can now both be SBTi-aligned while doing completely different things; a supplier having a target is no longer enough: you need to know what that target actually is.

Insets are still unresolved: The standard gestures toward the GHG Protocol to fill the definitional gap, but until that happens the boundaries of what counts as an inset remain unclear.

‍