Can I estimate business travel from a sample of employees instead of every trip?

Yes. The GHG Protocol's Category 6 guidance lets you extrapolate from a representative sample when collecting every trip is not feasible; its own example takes 400 of 4,000 employees. Group travellers with similar patterns, sample inside each group, scale up group by group, and write the method down.

When sampling is the right call

If one travel agent books every flight and train and sends a report with routes, use all of it; there is nothing to sample. Sampling earns its place when trips sit in reimbursement claims, cash vouchers and personal bookings, and pulling every one would take weeks. Many companies use full data for flights and a sample for local road and rail.

Group first, then sample

The guidance allows you to combine travellers with similar travel profiles. Do that before you pick anyone, because a random 10% of a 600-person plant can easily miss the five people who fly to Europe every quarter.

• Field sales on the road most weeks.
• Export and purchase staff who fly abroad: often small enough to cover everyone.
• Engineers visiting vendor and customer sites.
• Plant and office staff who travel a few times a year.

What to collect from each person sampled

• Kilometres for the year by mode: flights by distance band and class, rail, bus, taxi, own vehicle.
• Hotel nights, if you include them; they are optional in Category 6.
• The record behind each figure: ticket, claim or agent statement.

Keep these out of the sample

• Company-owned cars and vans: their fuel is Scope 1.
• Daily travel between home and work: Category 7, commuting.
• Trips a client pays for and books in its own name: they belong in the client's report.

Worked example: a 600-person company, kilometres before factors

• Field sales: 120 people, 24 sampled, 1,44,000 km by car and taxi; 1,44,000 / 24 = 6,000 km each; 120 x 6,000 = 7,20,000 km
• Export team: 15 people, all 15 covered from tickets: 3,10,000 passenger-km by air, no scaling needed
• Plant and office: 465 people, 47 sampled, 28,200 km by rail; 28,200 / 47 = 600 km each; 465 x 600 = 2,79,000 km
• 86 people sampled out of 600; multiply each mode's total by its per-km factor to get CO2e

Headcounts and distances are an example. Record each group's size, sample size, how people were chosen and the period covered.

What Senseible does here

Senseible reads the travel agent and hotel tax invoices you upload and keeps them in its Scope 3 supplier evidence trail, each with a SHA-256 fingerprint, so the records behind the sample stay on file. Scaling a sample to the whole company is a calculation you set and state.

• Flights from a travel agent invoice
• Hotel stays per night
• Employee mileage claims
• Field sales team travel
• Scope 3 supplier evidence

Sources: GHG Protocol, Technical Guidance for Calculating Scope 3 Emissions, Category 6 (data collection guidance, formula 6.1), GHG Protocol, Corporate Value Chain (Scope 3) Standard

Last reviewed 2026-10-31.