Why do two carbon tools give different Scope 3 totals for the same purchase data?

Two tools give different Scope 3 totals from the same purchases because they use different methods and factors: one may multiply rupee spend by an economic factor while the other multiplies tonnes by a material factor. Currency, GST, factor year, region and category mapping each move the result. Compare a sample of lines, not the totals.

The usual causes, largest first

• Method: spend-based (rupees x kg CO2e per unit of currency) against average-data (tonnes x kg CO2e per kg) against supplier-specific. On the same purchase these can differ by a factor of two or more.
• Factor database and region: a US or UK input-output factor applied to Indian purchases, against a multi-regional model such as EXIOBASE with an India region, against a material factor.
• Currency and price basis: the exchange rate used, the base year of the factor, and whether inflation since that year was removed.
• Tax: whether GST was stripped before applying a spend factor. Leaving 18% GST in raises a spend-based figure by 18%; the linked page on GST, inflation and currency covers the fix.
• Category mapping: the same line placed in Category 1 by one tool and Category 2 or 4 by another, or counted in two places.
• Boundary and gases: whether factors are cradle-to-gate, include transport, and use AR5 or AR6 global warming potentials.
• Duplicates: the same invoice uploaded twice, or a credit note not netted off.

How to reconcile two outputs

1. Export both at line level, not as category totals.
2. Pick the 20 largest lines by emissions in either tool.
3. For each line compare five things: quantity or spend, unit, factor value, factor unit and source, and category.
4. Mark which difference explains the gap: method, factor, currency, tax, category or duplicate.
5. Agree one rule for each cause and apply it to the full dataset.

Which number to trust

Neither total is right by default. Prefer the line that uses a physical quantity with a factor for that material and country over a spend-based line, and a supplier's own verified figure over both. A spend-based figure is still a fair estimate when it is the only data you have, as long as it is labelled as spend-based with wider uncertainty.

Worked example: one steel purchase, three results

• Purchase: 15 t of steel sections, Rs 9,00,000 before GST
• Tool A, spend-based, GST stripped: 9,00,000 / 88 Rs per USD = USD 10,227 x 0.8 kg CO2e per USD (illustrative) = 8,182 kg, about 8.2 t
• Tool A, GST left in: 10,62,000 / 88 = USD 12,068 x 0.8 = 9,654 kg, about 9.7 t (18% higher)
• Tool B, average-data: 15,000 kg x 2.0 kg CO2e per kg (illustrative) = 30,000 kg, 30 t
• Same purchase: 8.2 t, 9.7 t or 30 t depending on method and tax treatment

The exchange rate and both factors are illustrative, chosen only to show how the gap opens. Real factors depend on the database, year and region.

What Senseible does here

Senseible shows the quantity, factor and source behind every line, and labels spend-based estimates as spend-based with wider uncertainty, so its figures can be lined up against another tool's line by line. It does not audit or adjust another vendor's output.

• Spend-based vs activity-based Scope 3
• Spend-based Scope 3: GST, inflation, currency
• How to spot a wrong emission factor

Sources: GHG Protocol Technical Guidance for Calculating Scope 3 Emissions (2013), Category 1, UK DESNZ Greenhouse gas reporting: conversion factors for company reporting, US EPA Supply Chain Greenhouse Gas Emission Factors for US Industries and Commodities, v1.3

Last reviewed 2026-10-17.