When the carbon intensity page comes up in an investment committee session, the question that most often goes unasked is which portion of the figure on that page was counted and which portion was assumed. The presentation shows a single total, while the calculation beneath it carries, within the same cell, metered consumption data drawn from the organisation's own facilities alongside an estimate produced by multiplying the monetary value of purchased goods and services by a sectoral emission factor. In the same session, last year's page is placed beside this year's and the difference between them is read as improvement, although much of that difference may originate in an updated factor set or in a change in purchasing volume driven by price rather than by activity. The meeting proceeds to the next item without having drawn the distinction.
The same pattern is visible at the procurement desk. Asked to rank suppliers by carbon performance, a category manager typically holds one input — either the supplier's own declaration or an average intensity value assigned to the sector that supplier belongs to — with the consequence that, however large the real difference between two suppliers may be, the ranking the system produces tracks the ranking of spend almost exactly. The category manager recognises this and submits the report regardless, because what has been requested is not a defensible discrimination between suppliers but a completed table. This is not individual negligence; it is the natural outcome of an organisation demanding an answer in a domain where the answer does not yet exist.
The name for this pattern is the Scope-3 data gap — the absence of primary, verifiable data for indirect value chain emissions. What matters in the mechanism is not the gap itself but the manner in which the gap closes inside a decision process. A decision table does not tolerate a blank; it requires a number, and when a number is required a proxy is generated from economic input-output tables or sector averages. Everyone knows that proxy is an estimate when it is first produced; in the second year it is copied into the same cell, in the third year it becomes the denominator of a target declaration, and by the fourth year the person who knows who produced it under which assumption has left the organisation. An estimate begins to behave as a measurement precisely to the extent that the record of its provenance is not maintained.
This tendency is entirely functional under certain conditions, and ignoring that weakens the analysis. Across a supply base of several thousand vendors, the first-year cost of collecting primary data for every line comfortably exceeds, by a considerable multiple, the managerial value the inventory generates; a spend-based estimate is cheap and sufficient for establishing the rough order of magnitude and identifying where concentration lies. The problem does not reside in the shortcut but in the shortcut persisting after its purpose has changed. A figure produced for coarse mapping, once it becomes the measurement baseline for a reduction rate committed in a contract, is no longer an approximation but an unauditable obligation. The condition that validated the shortcut disappears silently while the shortcut remains in place.
The institutional consequence appears first in the purchasing decision. Within a spend-based inventory, replacing one supplier with another of materially lower emissions does not change the calculated figure at all, because the calculation is sensitive to the amount paid rather than to the identity of the counterparty. Under that structure the only routes to a lower calculated figure are buying less or buying the same goods more cheaply, and the second route frequently means shifting toward a supplier of higher carbon intensity. The link between the carbon target and the sourcing decision therefore does not merely break; it produces an incentive running in the opposite direction, such that an organisation may be reporting progress toward its target while doing the reverse across its value chain.
The second surface is customer contracts. Emissions clauses entering corporate supply agreements have been migrating steadily from declaration obligations toward performance obligations, with an annual intensity reduction committed and that commitment tied to a price tier or a renewal condition. At that point, if the measurement baseline supporting the commitment consists of average factors, the organisation carries penalty exposure on a figure it cannot move through its own performance. The same mechanism operates on the financing side: the margin ratchets of sustainability-linked facilities require a verifiable measurement baseline, and the first question a verifier asks concerns not whether the target is ambitious but whether the base year data is methodologically traceable.
The third surface emerges at a transaction desk. In the due diligence conducted on the transfer of a capital-intensive asset or an industrial group, the value chain inventory is no longer a compliance annex but a line item determining the scale of future compliance investment; the buyer prices not the inventory the target has declared but the cost of migrating that inventory to primary data and the extent to which the number may rise during migration. Replacing an average-based inventory with primary measurement usually pushes the total upward, since averages smooth concentrated high-intensity items. This uncertainty may not translate into a direct discount on the purchase price; more typically it settles into the timetable as a widened representation and warranty package, an elevated escrow ratio or a pre-closing methodology verification condition, and it delays completion.
The mechanism that neutralises this tendency is not collecting more data but recording the quality of data in the same place as the decision itself. The structure has four components. The first is data lineage: alongside every emission line, a separate field carries whether that line derives from primary measurement, supplier declaration, sector average or engineering estimate, and the total is never reported with that distinction erased. The second is a scope threshold: which spend band is migrated to primary data is determined not by emission magnitude but by the sensitivity of the decision to that line. The third is revision discipline: when the factor set or method changes, the base year is recalculated and the two effects — genuine reduction and methodology-driven movement — are shown separately. The fourth is embedding the data requirement into the contract; a data request that never becomes an annex to the supply agreement begins as a campaign and expires in its second year.
BEIREK's intervention in this area begins not with producing an inventory but with constructing the points at which the inventory attaches to decision processes. A mandatory field indicating the data quality of the line is placed into the investment approval form, the supplier prequalification file and the category-level sourcing decision; the field cannot be left blank, and a line resting on an average factor cannot proceed to approval on its own above a defined value threshold. On the same logic, the methodology record of the base year inventory is maintained as a separate document, with each revision annotated as to which line changed and why, so that a verifier arriving in the third year, or an acquirer's adviser, can audit the logic that produced the number rather than the number alone.
The second line of intervention sits on the contract and timetable side. Emissions commitments entering customer agreements are read before signature against the question of whether the measurement baseline underlying the commitment actually moves in response to the organisation's own decisions; where it does not, the commitment is either narrowed in scope or conditioned by a methodology transition clause. In parallel, the migration to primary data is structured not as a one-off project but as a sequence distributed across the supplier renewal calendar, so that each supplier assumes the data obligation together with its new contract and the inventory drifts toward primary data of its own accord within a three to four year renewal cycle. This spreads the cost across the existing commercial rhythm instead of demanding a large budget in a single year.
The mark of institutional maturity in this structure is not how low the inventory is but how precisely the organisation can state which portion of its own figure it trusts. An organisation that records its uncertainty may well report a higher number than one that dissolves uncertainty inside the total; the number produced by the second, however, does not hold at the first serious verification contact, and that contact tends to occur at the least convenient moment — in the middle of a financing close or a transfer process. What a decision table seeks is not certainty; it is advance knowledge of how much error tolerance sits behind each line.
The question worth asking, therefore, is not whether value chain emissions can be measured but whether the figure the organisation produces carries resolution sufficient to change a decision. Where a supplier substitution, a capital allocation or a product design choice leaves no trace in the inventory, that inventory is not a management instrument but a reporting obligation, and it carries the cost structure of the latter. The difference lies not in the quantity of data but in the record of where the data came from.