Can carbon accounting be fully automated, or what still needs a human?
No. Reading documents, classifying lines, applying published factors and checking arithmetic can be automated. Setting the boundary, choosing a base year, deciding which Scope 3 categories matter, accepting renewable energy claims and resolving flagged lines need a person who can be held to the decision.
What software can do reliably
- Read invoices, electricity bills, fuel slips and freight records into structured fields.
- Classify lines by HSN or tax code, and match fuels and electricity to a scope.
- Multiply activity by a published factor and record the factor's source.
- Recompute stored results, catch identical duplicate files and flag missing units or quantities.
- Produce the same output from the same input every time.
Decisions the GHG Protocol leaves to people
- Organisational boundary and consolidation approach: operational control, financial control or equity share.
- Base year and the policy for recalculating it after an acquisition, sale or method change.
- Which Scope 3 categories are relevant, and any exclusions with their reason.
- Market-based Scope 2: whether a REC, green tariff or power purchase agreement meets the Scope 2 Guidance quality criteria.
- Estimates where data is missing, and how they are labelled.
Work that stays manual in practice
Automation shrinks this list from every line to the exceptions. It does not remove it, and a tool that promises a finished number with no review is filling gaps with guesses.
- Lines flagged for missing units, unknown items or unusual quantities.
- Supplier data: asking for it, judging its quality, and following up.
- Documents the system cannot read clearly.
- Sign-off by a named person before the figures go to a buyer, lender or regulator.
Worked example: how many exceptions a batch can carry
- A batch has 50 lines; Senseible's Verified status needs at least 80% of lines to pass rule checks, with low risk
- 50 x 0.8 = 40 lines must pass, so up to 10 lines can fail and the batch still qualifies
- Those failed lines are still a person's job: fix, replace or exclude each one with a reason
What Senseible does here
Senseible automates reading, classification, factor application, duplicate detection and recomputation, and marks what it cannot decide for review. It does not set your boundary, base year or Scope 3 relevance, accept market instruments without your evidence, or replace your sign-off.
Sources: GHG Protocol Corporate Accounting and Reporting Standard, revised edition (2004), GHG Protocol Scope 2 Guidance (2015)
Last reviewed 2026-10-15.