Who’s Asking?

A fair question

The request normally arrives as a workbook. Fuel by month. Purchased electricity. Grid factor. Steam. Production volume. Scope 1. Scope 2. Sometimes Scope 3. Then the customer’s attributable share of all of it.

None of that is unreasonable, and it is worth saying so before saying anything else, because the argument that follows is easy to misread.

A company that sells a garment it did not manufacture typically carries most of its emissions outside its own walls. Its inventory is largely a description of other people’s factories. Its targets are, in practice, statements about other people’s boilers. Where the supply chain is opaque, the buyer reports an estimate built on an estimate, and a reduction it claims may be an artefact of the model rather than a change in the atmosphere. Spend-based accounting moves with price rather than with engineering: replace a coal boiler and the customer’s reported footprint may barely register it, raise prices and the same footprint worsens. Better supplier-specific data corrects that. It makes a downstream inventory responsive to what happens on a shop floor.

So the concession comes first, and it is not a small one. The buyer’s Scope 3 problem is real, it is technical, and it cannot be solved from a head office. Suppliers who treat every emissions request as an imposition are arguing against their own credibility. Manufacturing knows how to measure things. Measuring carbon is not the difficult part.

The difficult part begins later, and quietly.

Then the question changes

Somewhere in the later tabs, the workbook stops asking what happened and starts asking what will happen.

What is the plan for this facility next year. Which renewable-energy intervention will be installed, at what capacity. Which technology replaces the existing thermal system. What reduction does each intervention deliver. What is the approved budget. What is the timeline. What is the coal phase-out date.

Read individually, each is a fair question from a customer with a target to defend. Read together, they are no longer disclosure. The first set asks a supplier to report. The second asks a supplier to set out how it intends to run and refinance its asset base.

Three different things have been folded into one form here, and they are worth separating. Measurement asks what happened. Visibility asks how the footprint will decline, and there is a reasonable case for it, because a customer carrying a target cannot manage what it cannot see coming. Influence is the third question, and the one that usually goes unnamed: how far into the industrial choices should the requirement reach, which asset, which technology, which year, which budget line.

Figure 1 · Measurement, Visibility, Influence

The request changes category before the form says so.

Request Escalation

Three-stage continuum separating measurement, visibility and influence in buyer emissions requests.

01

Disclosure

Measurement

What did you emit?

02

Forward view

Visibility

How will this decline?

03

Capital reach

Influence

Which asset, which technology, which year?

reporting

planning

The escalation is procedural rather than declared. Little in the process marks the point at which the request moves from measurement to direction. The form simply gets longer.

Measurement is legitimate. Prescription is different.

What the frameworks actually ask

It is tempting for a supplier to answer this by pointing at the rulebook, and equally tempting for a buyer. Both should be careful, because the rulebook is narrower than either side tends to assume.

The major climate accounting, disclosure and target-setting frameworks govern the reporting company. They set out how a corporate inventory is built, what must be disclosed, and what a credible target looks like. They are demanding documents. They are also, on the specific question at issue here, quiet.

The GHG Protocol’s Category 1 guidance recognises supplier-specific, hybrid, average-data and spend-based calculation methods. Supplier-specific primary data is encouraged where it improves the inventory; it is not universally compulsory. The European reporting standards permit value-chain estimates, including sector averages and proxies, where direct information cannot be obtained after reasonable effort. IFRS S2 prioritises better data while recognising secondary information and applying a proportionality test. California’s climate disclosure statute expressly allows both primary and secondary data, including industry averages.

Target-setting works the same way. A supplier engagement target commits a buyer to bring a defined share of its suppliers into science-based target setting, at minimum covering the supplier’s own Scope 1 and 2. It does not, as a condition of target validation, require the buyer to collect facility-level inventories, buyer-specific allocations, equipment retrofit schedules or capital expenditure plans from those suppliers. Those are not the requirement. They are one way of pursuing it.

None of which makes the facility questionnaire illegitimate. Industry tools go further deliberately, and openly. The apparel sector’s own facility assessment module asks sites for improvement opportunities, cost and benefit, approved budgets and implementation timelines. That is a design choice made for defensible reasons: a target with no implementation behind it is a press release, and buyers have seen supplier plans that consisted of a signature.

The point is where the obligation ends and the choice begins. The framework creates the climate obligation. The buyer decides how far that obligation travels into the supplier’s operating model. That decision is not neutral, and no framework reviews it.

Your reporting unit is not automatically my investment unit

This is the centre of the problem, and it is not a moral one.

A facility is a useful boundary. It is where the meter sits, where the fuel burns, where the permit is issued, where the emissions occur. For measurement, allocation and accountability the site is often the right frame.

But the boundary that is useful for measuring emissions is not automatically the boundary at which a business decides where capital goes.

Consider what a multi-site industrial group actually owns. Not a row of independent factories, but a portfolio: assets of different ages, shared utility infrastructure, sites inside industrial estates with common services, divisions with different product economics, a spinning unit whose steam profile has nothing in common with a garment unit next door, and one balance sheet from which all of it must be funded. Capital in that structure is rarely allocated by address alone. It is allocated by return, by feasibility, by asset cycle, by risk, and by what the group can finance in a given year.

Reporting precision does not create investment logic. A customer can require, correctly, that emissions be reported per site. It does not follow that the supplier should optimise per site. Attribution and optimisation are different problems. Carbon reduction wants the largest abatement per unit of capital, wherever that abatement sits. Customer reporting wants the number attached to the correct factory. Those often point the same way. Not always.

A factory can be a valid emissions boundary without automatically being the right unit at which to optimise an entire transition. That is an analytical distinction rather than a clause in any standard. It is also how industrial capital is normally allocated.

The counterweight matters here. Portfolio logic can be abused. A group that can point to a better use for capital elsewhere every year is a group whose worst site does not move, and facility-level questions are one of the few instruments that expose exactly that. Some interventions are irreducibly site-specific: a boiler, a permit, an effluent system, a building’s structure, a local regulation. Where a customer insists on a plan for the site that actually makes its goods, that insistence can be what finally moves a neglected asset. The argument is not that the facility is the wrong unit. It is that the facility is not automatically the only unit.

Figure 2 · Reporting Unit vs Investment Unit

Allocation is not optimisation.

Boundary Test

Two-panel diagram comparing a buyer Scope 3 reporting map by facility with a supplier capital map based on shared assets, divisions, infrastructure and project sequence.

Buyer reporting map

Facility A

Facility B

Facility C

Emissions are measured, allocated and attributed by site.

Allocation
is not
optimisation

Supplier capital map

01

Shared utilities

02

Energy asset

03

Division economics

04

Project queue

Capital moves through feasibility, asset cycles, risk and financing capacity.

Carbon is not impressed by symmetry

The engineering makes the case more plainly than the accounting does.

Two sites in the same group, producing to the same certification, can have little in common physically. One roof can carry an array and has enough service life left to justify one. The next cannot, or will need replacing well before the panels reach the end of theirs. One site has land. One sits on a leased plot where the roof belongs to the landlord.

Grid conditions differ. Process heat differs more, and process heat is where the difficulty sits. Whether a heat pump is viable at a given site depends on the temperature the process requires, the temperature of the available heat source, the electricity tariff and how the equipment integrates with the existing steam system. Electrification is not automatically a reduction either. Where the grid is carbon-intensive relative to the fuel being displaced, an electric boiler can raise combined emissions rather than lower them, which means the same intervention can be progress at one site and regression at another in the same country.

None of this is an excuse. It is why a rational transition looks uneven. Transition planning can be asymmetric because physical conditions are asymmetric, and a template that expects the same class of intervention at every site in the same year is asking about paperwork rather than about carbon.

The rational tonne is not always the neatest tonne.

The same asymmetry runs through time. In practice, transition is sequenced. Lower-cost efficiency measures often move first, and the savings help fund what follows. Renewable electricity where the site and the grid allow it. Heat recovery. More capital-intensive changes follow as technology, infrastructure, asset cycles and finance allow, and major thermal work is often timed to an asset’s replacement rather than forced into the middle of its life.

Some of the largest industrial decarbonisers describe the same pattern in public. ArcelorMittal sets out a sequential rollout of electric arc furnace capacity and has deferred some direct-reduction pathways where clean power, hydrogen or policy economics are not yet in place. Dow took a final investment decision on a concentrated low-carbon investment at a single Canadian site, where feedstock, carbon storage infrastructure and public incentives converged, then delayed construction as market conditions changed. Neither is an evasion. Both are what capital-intensive transition looks like.

The pressure a supplier feels is rarely one customer demanding everything at once. It is the aggregate. Several customers, each entitled to ask, each asking about the sites that make their goods, each on its own reporting calendar. No single request is unreasonable. The sum can convert a sequenced plan into an expectation of concurrent visible progress across a portfolio that was never going to be funded that way in one year.

Transition plans are usually sequential. Buyer questionnaires can quietly make them concurrent.

Figure 3 · Sequential Transition vs Concurrent Visibility

One capital queue meets several reporting calendars.

Timing Conflict

Supplier transition activity is sequenced across five years, while three customers each request visibility across multiple supplier sites within the same reporting cycle, creating concurrent reporting pressure.

Supplier transition sequence

Year 1

Efficiency

Year 2

Renewable electricity

Year 3

Heat recovery

Year 4

Thermal conversion

Year 5

Residuals

Buyer visibility window

Customer A

Plan visibility: now

Site 01 Site 02 Site 03

Same reporting cycle · multiple sites

Customer B

Plan visibility: now

Site 01 Site 02 Site 03

Same reporting cycle · multiple sites

Customer C

Plan visibility: now

Site 01 Site 02 Site 03

Same reporting cycle · multiple sites

The capital queue is finite. A project moved up the order because a customer’s template asks for it this cycle can displace one that would have delivered more abatement per unit of capital.

The industry has run a version of this experiment before. Social compliance began as a legitimate demand for assurance and became a proprietary format per customer, with the audited party paying for most of the duplication and rarely being funded to do it. That is social-audit evidence rather than climate-reporting evidence and should not be relabelled as the latter. The architecture, though, resembles the one now being assembled under a different heading.

One boiler, several targets

A supplier facility often serves more than one customer. When it installs a thermal system, contracts renewable power or rebuilds a utility loop, the emissions intensity of the goods it produces can fall. Where each customer’s method is sensitive to supplier-specific performance, one intervention can improve the reported trajectory of several buyers.

This is not an accusation of double counting. Whether a given buyer’s inventory captures the reduction depends on its method and its claims, and that is an accounting question with an accounting answer. The commercial question sits alongside it and is simpler.

One party financed the asset, carries the debt, absorbs the execution risk and the downtime, and holds whatever the asset is worth if the technology dates or the customer moves on. Several customers may see the benefit reflected in their reported trajectory.

That is tolerable while the request is disclosure. It becomes harder to justify as the request deepens. How many customers can expect visibility into the same action plan, and influence over its sequence, before the investment relationship itself becomes part of the conversation?

Figure 4 · One Investment, Several Scope 3 Relationships

The asset is financed once. The reported benefit can travel through several customer relationships.

Commercial Exposure

Hub-and-spoke diagram with one supplier facility intervention connected to three buyer Scope 3 relationships.

Buyer A

Scope 3 relationship

Buyer B

Scope 3 relationship

Supplier facility

One thermal or utility intervention

Debt, execution risk, downtime and asset life sit here.

Buyer C

Scope 3 relationship

Accounting note

Recognition depends on each buyer's method and claims.

Commercial time and climate time

A transition plan runs across years. So does the asset. The commercial relationship supporting both may not.

That needs stating carefully. Not every sourcing relationship is short. Many are long and stable in everything except paper. Some buyers place volume with the same manufacturer for a decade and forecast well enough to plan against.

But long-standing is not the same as committed. The industry’s own purchasing-practice benchmarking puts planning, forecasting and order placement among its weaker areas, which is precisely where investment planning needs strength. Manufacturer groups have proposed, without any expectation that it binds anyone, that buyers pay for reserved capacity they do not use. Producers do not propose that when demand visibility is abundant.

So the question is bankability. What makes a multi-year facility transition commitment rational when the commercial relationship supporting it is not equally durable? A customer requirement does not by itself underwrite the cash flow of a long-lived asset.

Suppliers deserve their share of this scrutiny. Some underinvest. Some delay until a requirement becomes unavoidable. Some overstate barriers, and some ask customers to fund projects whose internal returns were already sufficient. Lower energy cost, lower water cost, resilience against fuel volatility, process stability and market access mean that a good number of decarbonisation projects pay for themselves without buyer funding, and a supplier who pretends otherwise is doing what it accuses the buyer of doing.

Nor is it true that buyers do nothing. Multi-brand vehicles now exist specifically to lower the cost of capital for factory-level projects, using first-loss capital alongside commercial bank lending, and blended-finance platforms backed by development finance institutions are being structured for this sector. They are real, and they address the right problem. What they are not yet is normal. Reviews of large fashion companies still identify only a handful evidencing direct supplier project financing, against sector models that put cumulative financing for the industry’s decarbonisation somewhere in the region of a trillion dollars. Those are models rather than invoices. They do not describe a gap that supplier balance sheets absorb quietly.

A plan has two sides

The alternative is not silence from suppliers. It is a plan with commitments on both sides.

The supplier side of that plan is not light. A credible baseline, measured rather than asserted. Facility data at the granularity the customer’s inventory actually needs. Technical feasibility work, honestly done, including the projects that do not work and why. A sequenced pathway with the reasoning behind the sequence exposed rather than hidden. Costs. Expected emissions outcomes. Named implementation responsibility, and dates the supplier is prepared to own. A supplier that cannot produce that has not earned the conversation about reciprocity.

The buyer side is where the architecture is currently thin. It need not be a cheque, and reducing it to price is the fastest way to make the argument sound self-serving. It can be demand visibility extending past the current season. Sourcing tenure. Preferred allocation. Access to cheaper capital through a guarantee or a first-loss layer that costs the buyer far less than the project itself. Co-investment where the intervention is specific to that buyer’s requirement. Technical support. Offtake for the lower-carbon output. Faster payment. Or the least expensive item on the list and often the last to be offered: a common data format, so that the same transition is not rewritten in six templates a year and the hours saved go into engineering rather than translation.

None of that says every buyer must provide all of it, or any particular item in a particular case. It says that where a supplier’s pathway matters enough to a customer to require detailed forward commitments, the relationship should at least ask what credible commitment belongs on the other side. Any of those instruments makes a pathway more bankable. Most of them cost less than the alternative, which is a supply base that answers every questionnaire and installs very little.

Figure 5 · The Two-Sided Transition Plan

A plan becomes credible when both sides carry obligations.

Shared Plan

Split contract-style figure showing supplier commitments on one side and buyer commitments on the other, joined by a shared transition plan.

Supplier side

Baseline

Feasibility

Sequencing

Cost

Implementation

Data

Shared
transition
plan

Buyer side

Demand visibility

Tenure

Finance and guarantees

Allocation preference

Data harmonisation

Risk sharing

Who’s asking?

Return to the workbook. The supplier will complete it. The data will be assembled, the intensity calculated, the buyer’s share allocated, and the file uploaded before the deadline, because that is what a working supply relationship does and because the underlying need is real.

But something should be said out loud at the point where the tabs stop describing the past.

A request for information is information gathering. A request that names the technology, the site, the asset to be retired, the date and the budget line begins to shape a decision set that belongs to someone else. Not control. Influence, of a kind that grows with commercial leverage and with the specificity of the ask. A buyer that names the outcome leaves the supplier its engineering judgement. A buyer that names the means has started to write the supplier’s capital plan without appearing on the supplier’s balance sheet.

Who’s asking is not a challenge to the right to ask. It is a question about what the answer should mean for the relationship.

If a supplier’s transition pathway is material enough to a buyer’s climate strategy that the buyer needs to see it, sequence it and verify it, then it is material enough to be planned together. And a plan with obligations on only one side is not a plan. It is a submission with a deadline.