Methodology.

Scope & boundary.

cobank reports company-level emissions under the GHG Protocol Corporate Standard. The organisational boundary is set by operational control: an entity appears in the ledger if the reporting company has the authority to direct its operating policies, regardless of equity share. This is stated once per workspace and held constant across periods unless the company formally restructures.

Within that boundary, cobank calculates Scope 1 (direct combustion and fugitive emissions), Scope 2 (purchased electricity, steam, heat and cooling, reported both location-based and market-based where residual-mix factors are available), and the material portion of Scope 3 under the GHG Protocol Scope 3 Standard: Category 1 (purchased goods and services), Category 3 (fuel- and energy-related activities), Category 4 (upstream transport), Category 6 (business travel), and the material sub-portions of Categories 2, 5 and 7.

What cobank does not report on by default: Scope 3 Categories 8 through 15 (use-phase, end-of-life, franchises, investments, downstream transport and leased assets) are out of scope unless the workspace explicitly configures them, because the data quality of category-wide spend-based estimates at that end of the value chain is materially lower than anything else in the ledger. Non-energy Scope 1 sources for which no published factor is available are flagged rather than estimated. Silence on a category is preferable to a number the auditor cannot defend.

Calculation methods.

Every kgCO₂e figure in cobank is produced by one of three methods, always tried in this order and recorded on the line.

Surcharge-zero rule. On transport and energy invoices, freight surcharges (BAF, CAF, LSS, fuel surcharges, demurrage), taxes (VAT, MwSt, TVA), administrative fees, service charges, and retrospective adjustments are assigned activity_role of surcharge / tax / fee / adjustment in the ledger and emit zero kgCO₂e. This is a direct application of GHG Protocol guidance on avoiding double-counting: the physical activity that generated the emission is already captured on the parent primary line, and billing the same tonne of CO₂e twice because the carrier split the freight rate across three invoice rows would inflate the total. The classifier is deterministic, not probabilistic: a line matches the rule or it does not, and the ledger records the rationale and the parent line it defers to.

Factor sources.

cobank uses a curated library of published emission factors. Every factor carries a source, a version, a region of applicability and a validity window. When a line item is calculated, the factor version in force on the invoice date is pinned to the line via factor_version_id and held there permanently.

Selection hierarchy. For every line, cobank applies the curated calculation library (over 75 factors across 14 source labels) in a fixed order. For grid electricity the factor is the one published for the country of consumption: BAFU for Switzerland, the Umweltbundesamt for Germany, the European Environment Agency for the other EU member states, and the EU-27 average as a disclosed fallback where no national factor is held. Otherwise: DEFRA 2024 (for freight, fuel combustion and business travel) → GLEC 2025 / IMO 2025 (for logistics chains) → EEIO spend-based factors as the last resort. A separate semantic pool of about 1’400 DEFRA 2024 factors (pgvector-indexed) produces alternative suggestions on the Provenance Card at a 0.30 cosine-similarity threshold with a required subcategory match: it never drives a calculated number. A line with no curated match is routed to the review queue, not silently approximated.

Optional add-ons. Ecoinvent is available as a customer-procured add-on on enterprise plans when a customer's audit firm or internal methodology specifically requires it; cobank does not redistribute ecoinvent by default because it is not required for credible Swiss or EU voluntary reporting. A customer-supplied factor library (override mode) is also available on enterprise plans; see enterprise plan.

Version log. Factor versions are pinned per line at insertion time via factor_version_id. When a source publishes a new version, prior reported periods are not automatically restated; historical numbers remain bound to the factors that were in force on the invoice date. Restatements are documented per-period on request, via a period-snapshot hash that records the exact factor set in use at close.

Attribution. DEFRA / DESNZ conversion factors contain public sector information licensed under the Open Government Licence v3.0. German grid values: Quelle Umweltbundesamt, CLIMATE CHANGE 16/2026, converted from g/kWh to kg/kWh. European country grid values: © European Environment Agency, re-used under the EEA re-use policy (CC BY), converted from g/kWh to kg/kWh. Swiss values: Bundesamt für Umwelt BAFU, re-used with BAFU’s written confirmation of 4 May 2026. No market-based residual mix is applied for EU supply: the AIB European Residual Mix carries no re-use permission, so where no supplier-specific factor or cancelled instrument exists cobank applies the consumption-country grid average and says so on the line, as the GHG Protocol Scope 2 Guidance permits when a residual mix is unavailable.

Confidence & data quality.

Every line item carries a confidence score between 0.00 and 1.00 that summarises how much the calculation can be trusted. Three inputs feed the score: the quality of the underlying factor source (physical-method factors from BAFU or DEFRA score higher than spend-based EEIO fallbacks), the certainty of the classification (a strong vendor match and an unambiguous description score higher than a generic line on an unfamiliar supplier), and a missing-data penalty for each field that had to be inferred rather than read from the invoice. The score is a review signal, not a correction factor: a low score does not reduce the reported kgCO₂e, it determines where the line goes in the workflow.

Confidence is visible on every line of the ledger and on every row of the Provenance Card. When a reviewer overrides a classification, the override is recorded alongside the original suggestion, the factor library rows it was drawn from, and the identity of the reviewer; the confidence score is recomputed rather than overwritten.

Immutability & audit trail.

Every write to a line item (creation, edit, override, review, close) passes through a hash chain. Each row records a row_hash over its own fields and a prev_hash pointing at the previous state. UPDATE and DELETE are blocked by row-level-security policy on any line whose reporting period has been closed, so a closed period cannot be silently rewritten. Amendments open a new version with a visible restatement trail rather than quietly mutating history. Every factor version is pinned at write time, so historical calculations remain reproducible even after the factor library advances.

This means every number in a cobank report can be reconstructed from its inputs: the invoice it came from, the line it was extracted into, the factor that was applied, and the identity of whoever signed off on the final classification. Reproducibility is not a marketing claim here: it is the shape of the database.

How cobank is operated.

A statement is only as strong as the process behind it. cobank runs one loop per reporting period.

Where the rule lives. The amendment trail is enforced by the database, not only by the application. A sealed line refuses a write that does not arrive through the amendment path, and so does the invoice behind it — its date, its supplier and whether it has been archived are part of the sealed evidence. The application asks for a reason; the database is what makes the asking unavoidable.

Verifiability controls.

Zeros are never ambiguous: every line resolves to one of five states (measured · instrument-backed zero · rule-based zero · not measured · zero, unexplained), and the data-quality summary reconciles to the line count. Coverage is reported on two labelled bases, by line count and by spend.

Where a line's calculation basis is stored, the report build re-performs the arithmetic and surfaces any difference as an exception; where it is not, the line is disclosed as not re-performable from stored fields — routed to source documents rather than silently passed. Invoice totals are reconciled against the sum of extracted lines. Duplicate candidates are flagged for judgment, never silently merged.

What the seal covers, stated precisely: a sealed period's lines cannot change, and neither can the documents behind them. A document dated into a closed quarter that arrives after the close is held unstamped and reported as an addition — it is not folded silently into a sealed total, and it does not move one.

Last reviewed: 7 August 2026.

Status of independent assurance.

For the avoidance of doubt: the reports that leave a cobank workspace are internal management figures until an accredited assurance provider has tested them. Where a customer wants limited or reasonable assurance under ISSA 5000, the workbench supports engagement-scoped read access, risk-based sampling, per-assertion workpapers and a signed evidence package that the firm can rely on. The assurance opinion remains the firm's; cobank provides the working paper underneath.

See also: What VSME actually requires of a Swiss SME, the primary-source walk-through for a Swiss CFO whose EU customer just asked for VSME-aligned data.

Revision log.