This is a support sheet for activity data, not a carbon calculator. Record journey legs in kilometres. Say whether each number is the vehicle's distance, one person's distance, or an already aggregated passenger distance. Count a shared vehicle once. Leave people who did not reply out of the aggregate. Extrapolation stays off. The absence of name columns does not promise anonymity or GDPR compliance.

The carbon guide is where a later inventory would name a conversion-factor year and a boundary. This file has no emission column. Staff travel policy stays on the business travel guide. Do not describe unanswered journeys as a measured footprint. That claim belongs with the green claims guide.

Decide what each number means

Use kilometres for the whole file. Do not mix miles into a label that says kilometres.

Each distance is one of three things:

  • vehicle distance: the kilometres the vehicle travelled
  • per-person distance: the kilometres one person travelled on that leg
  • passenger distance: a figure that is already person-kilometres, not a single person’s trip

Multiply only in this one case. If you start from a per-person distance and you want passenger distance, multiply that person’s kilometres by the number of people who travelled that same leg, once. Record the product as passenger distance, or keep the per-person row, but do not keep both in the same total. Do not multiply vehicle distance by the number of passengers. A car that travels 20 kilometres outward with two people is 20 kilometres of vehicle distance for that direction, not 40. The two passengers are not a second 20 kilometres.

Keep the three bases in separate totals. This sheet does not add a vehicle total to a per-person total.

Count people, legs and shared cars separately

response_count is the number of individual people who replied. It is not the number of legs and not the number of passengers. It cannot be higher than the number invited. Both counts are whole numbers, zero or greater. Distances are finite and zero or greater. A blank distance is unknown, not zero.

The number of people still missing is the invited count minus the response count. Do not subtract the replies from the response count, and do not invent journeys for the people who did not reply.

To spot a shared car without collecting names, phone numbers, email addresses or registrations, agree a short code before travel, such as red-2, and ask each person in that car to use it. If two rows share the code, the direction and the distance, they are one vehicle. Count that vehicle once in the vehicle total. If you cannot tell whether two car rows are the same vehicle, mark the duplicate unresolved and leave it out of the verified aggregate.

A return is either a separate journey leg or a note that the return repeats the outward leg. repeats_outward means the return happened and was the same distance. Count that return once, in the same vehicle total, as well as the outward leg. It is not a reason to leave the return out, and it is not multiplied by the number of passengers. Keep an outward-only subtotal separate from that round-trip total. Do not also enter a return row for the same vehicle. If the marker and an explicit return row are both present, leave that vehicle out of the verified aggregate. Do not add the implied return on top of the explicit row.

Extrapolation stays off

The extrapolation cell stays not_applied throughout this sheet. Writing an assumption in the note does not turn extrapolation on. This resource does not contain a method that has been validated for scaling respondents up to the invited count.

Leaving out name and contact columns follows the data-minimisation idea on the ICO’s page. That guidance is marked under review because of the Data (Use and Access) Act. Using this sheet is not a finding that a survey is anonymous or that it complies with UK GDPR.

Fill in the survey

Download both files. Save a copy of the blank file and type into that copy. You type every cell. The spreadsheet does not multiply passengers or scale the missing replies.

Write one header, then one row per journey leg. response_id is a serial such as R1, not a name.

Worked example: an excerpt, not the whole survey

Northgate Studio’s fictional morning invited 40 people and received 10 replies. Missing people are 40 minus 10, which is 30. The example does not invent their journeys. The leg rows are an excerpt. They are not all 10 replies, so they are not a survey total.

Leg What it is What you can add
R1, car, 20 km outward, two passengers, code red-2 Vehicle distance. The return repeats this 20 km and is not given its own row. 20 km in the outward-only subtotal. 40 km in the round-trip vehicle total, once each way. Not 80 km.
R2, same code, same outward leg, same 20 km The same shared car. Nothing further. It is already counted.
R3, rail, 15 km out and 15 km back Two per-person legs. The return is its own row. 30 km in the per-person total only. Not added to the 40 km.
R4, car, 18 km, no reliable code You cannot tell if another row is this vehicle. Excluded from the verified aggregate.
R5, car, code blue-1, 12 km out marked as a repeated return, and a separate 12 km return row The return is written twice. Excluded. Do not add 36 km, and do not silently keep both.

The outward-only vehicle subtotal in this excerpt is 20 kilometres. The represented round trip for that same shared car is 40 kilometres, counted once in each direction. The per-person rail figure is 30 kilometres and stays in its own total. Extrapolation is not applied, and the assumption cell is blank.

Limits

Accommodation, crew travel and streaming are outside this file. The 2026 conversion-factor collection, updated 11 June 2026, is for a later calculation. No factor value is copied here.

Sources