How do you survey attendee travel and report carbon estimates honestly?

You've paid for a shuttle, rewritten the travel page and encouraged people to take the train. Now a sponsor wants to know how much carbon your event saved. You open the attendee survey and find a mixture of outward journeys, vague distances and people saying they meant to cycle. What can you actually report?

A useful event attendee travel survey records the journeys people made, including the way home, and keeps those answers separate from assumptions about everyone who didn't reply. Collect transport types, distances and shared-car occupancy in a consistent format. Then report who answered, what you could calculate and where the estimate still depends on missing information.

You don't need to wait for a perfect audience census before learning something useful. You do need to resist turning a tidy percentage into a claim the survey cannot support. A few detailed answers might explain why your shuttle isn't working. They won't automatically tell you the emissions of every person who walked through the door.

This guide walks through choosing the survey's purpose, building the questions, reaching attendees and checking the results. There's a questionnaire you can adapt, a worked example of incomplete responses, and wording for sharing findings with your team or sponsors without making them sound more certain than they are.

Decide what the travel survey needs to help you change

Start with a decision you can make. Perhaps you're deciding whether to keep the station shuttle, whether a different finishing time would help rail passengers, or whether parking demand justifies a costly field rental. Those questions need slightly different information. Asking everything you can think of creates a longer form without necessarily improving the decision.

For a shuttle review, you want the connections people used, the return they needed and what stopped others using it. For a venue comparison, broad starting areas and complete journeys matter more. For a carbon estimate, whoever calculates it needs transport details that match their chosen method. Agree those requirements before sending the survey.

Culture for Climate Scotland's guide to audience travel recommends collecting mode and distance, considering different outward and return journeys, and reviewing the calculation tool before writing the questions. That's a sensible order. You'll struggle to repair a form after discovering that your calculator needs information you never asked for.

Keep the practical decision visible even if you're preparing an environmental report. The survey should help you understand whether people could reach your paid event and get home at an acceptable cost. Someone who drives because the last bus leaves before the finale has told you something about the programme, not simply their attitude towards sustainability.

Our guide to sustainable event travel covers designing those alternatives. Here, the job is to collect enough information to judge what happened and choose the next improvement. If the form won't change a decision, narrow it before asking your audience to spend time completing it.

Set the boundary before counting journeys

Write a short internal description of whose travel you're studying and which journeys belong in it. For a single evening show, you might include individual audience members' journeys to the venue and their journeys home or back to overnight accommodation. Keep crew, performers, suppliers and exhibitors in separate records if their travel is also being measured.

Decide whether complimentary guests are included. They still travel, even though they didn't buy a ticket. If your total describes all audience travel, excluding them because the sales report contains only paid admissions leaves a gap. Name the audience consistently in the survey, calculation and final report.

A destination conference needs more thought than a local show. Someone may fly into the city, travel between a hotel and the venue, and remain for a holiday afterwards. Ask enough to distinguish those activities. Agree how the event-related share of a combined trip will be estimated, and disclose that rule rather than quietly including or excluding the entire holiday.

For a multiday event, separate unique people from attendance days. One person attending twice may make two return journeys. Another may camp on site and make one arrival and one departure. A count of ticket scans cannot describe both patterns without more information, and re-entering after lunch doesn't create a new person.

Choose the unit of response too. The questionnaire below collects one person's travel. A purchaser who bought four tickets answers for their own journey and shares the invitation with their guests. Don't multiply that purchaser's journey by four unless you have explicitly collected a group response and verified who it covers.

Group reporting can work, but it needs a different design. You must know which people the answer covers, whether they travelled together throughout, and whether any also replied individually. For a routine first survey, individual responses are easier to explain and check. Make that choice explicit in the opening instructions.

Ask about completed travel, and keep plans separate

Before the event, travel intentions can help you choose services. After the event, you want the journeys that happened. Keep those two records separate. Someone who selected rail at checkout may have driven after a cancellation. Their earlier answer is useful planning information, but it cannot quietly become their completed journey.

For an evening event, send the main travel request after people have had a reasonable opportunity to return. For a residential event, allow for later departures. State the journey period you're asking about and offer a way to say the return is still unfinished. Follow up on that record if your process allows corrections.

An exit survey can collect the outward journey accurately enough for some purposes, but the person may only know their intended return. Mark it as intended. If the last train is cancelled later, an outward-plus-planned-return record will miss the taxi that actually took them home. Don't present both parts as completed travel because they appear on the same form.

You can use a short on-site question to identify a problem while there's still time to help. Keep that support route separate from the later analysis. A traveller asking about a missed connection needs practical assistance, and their request shouldn't depend on completing a carbon questionnaire or agreeing to future marketing.

Set the collection dates in advance. Record when the invitation and any reminder went out, when replies closed, and what changed during collection. You can then explain whether people answered on the journey home or several days later, rather than treating all recollections as though they were collected under identical conditions.

A travel questionnaire you can adapt

Use this as a starting form for individual adult attendees. Adapt the dates, location, journey boundary and answer choices to your event and calculation method. If children are part of the audience, plan an appropriate parent or guardian reporting approach rather than assuming the same invitation works for everyone.

The introduction might read: "We'd like to understand how you travelled to and from this event so we can improve future travel arrangements and prepare an estimated travel footprint. Please answer for yourself, including every part of your journey. Participation is optional. It's fine to choose 'not sure' rather than guess. Our privacy information explains how we'll use and keep your answers."

Add the actual closing date and the correct privacy link before sending. Don't promise a completion time until you've tested the form. Someone walking from a nearby flat may finish quickly; someone describing flights and several connections will need longer. Branching questions let each person see the detail relevant to their journey.

Confirm attendance and the journey period

Ask: "Did you attend this event?" Offer yes, no, and an appropriate route for someone unsure which booking the message concerns. Route nonattendees away from the completed attendee-travel calculation. If understanding nonattendance matters, offer a separate optional question without treating that answer as a journey to the venue.

Ask: "Which event date or dates did you attend?" Use the actual dates. For a single show, display the date in the introduction instead of asking people to enter it again. For repeated visits, explain whether they should describe each visit separately or identify a journey pattern that genuinely repeated.

Ask: "Have you finished your return journey?" Offer yes, no, and not applicable with an explanation box. A person staying in the city may have completed the local trip back to accommodation while their longer journey remains unfinished. Your instructions need to tell them which boundary applies.

Record the outward journey in parts

Ask: "Where did the journey to the event begin?" Offer a broad town or area, plus whether that was home, work, overnight accommodation or another starting point. Ask for a country when international travel matters. Use the least location detail that still lets you analyse the journey sensibly.

Ask: "Which types of transport did you use to reach the event?" Let people select every type used. Relevant choices may include walking or wheeling, bicycle, private car, taxi or ride-hail, local bus, train, event shuttle, coach, ferry and air travel. Include other with a description, plus not sure where necessary.

For every selected type, ask: "About how far did you travel using this transport on the way to the event?" Make the unit explicit, with a clear choice of miles or kilometres if you accept both. Explain that this asks for the distance in that part of the journey, rather than the whole journey repeated beside every transport type.

Allow a route description or not sure when someone cannot estimate distance. You may later estimate it using an agreed routing method, but preserve the difference between a respondent's distance and an analyst's estimate. If you use distance bands, choose them with the analyst and record how each band will be converted for calculation.

Add the details needed for the selected transport

For a private car, ask how many people were in the vehicle, including the driver, for that part of the journey. Ask vehicle or fuel type only to the level your chosen method needs, and allow not sure. Occupancy can change when somebody is collected along the route, so offer another journey part when that would materially affect the estimate.

For a taxi or ride-hail journey, ask how many passengers shared the ride, including the respondent. Record private lifts using the car questions. Distinguish the attendee's occupied journey from additional vehicle travel the attendee may not know. Don't make them invent how far a taxi drove before collecting them. Flag any extra-distance assumption for the analyst, using the selected method's rules and avoiding unsupported precision.

For flights, ask the relevant airports and cabin class if the method requires it, with not sure available. Distinguish connections so a transfer airport doesn't disappear inside one broad departure location. For a ferry, identify the route and whether the attendee travelled with a vehicle if that distinction affects the calculation.

For an event shuttle, collect the route or pickup point and direction used. You can compare that with boarding records and the operator's vehicle data later. A respondent shouldn't need to know the coach's engine details or empty return mileage. Ask the supplier for the operational records your calculation needs.

Describe the return without forcing a second identical form

Ask: "Was your return journey the same route, distance and transport as your outward journey?" Include occupancy in that instruction where shared vehicles were used. If yes, explain that the outward journey will be repeated once for the return. If no, show the journey-part questions again for the actual return.

Ask: "Where did the return journey end?" Use the same level of location detail as the outward question. A person leaving work for the show and returning home may use the same train but travel different distances. A simple transport checkbox would miss that difference.

Include an optional question for additional visits or local journeys within your agreed boundary. A weekend pass doesn't tell you whether someone commuted each day. Give a clear example of the journeys to include, without asking people to reconstruct unrelated meals, sightseeing or errands you have decided to exclude.

Finish with questions you can act on

Ask: "What, if anything, made your journey difficult?" Offer an optional answer in the person's own words. If you need structured choices, include the barriers you can act on, such as unsuitable departure times, missing connections, cost, unclear information or access arrangements. Allow more than one reason and something else.

Then ask: "Is there a change we could make that would help you travel differently next time?" Keep this optional and describe it as feedback. Someone saying they might use a shuttle has expressed interest, not made a reservation. Don't use these answers alone to guarantee a full coach or forecast a carbon saving.

Test the form with people who made different journeys. Include a straightforward local journey and a more complicated one involving several transport types. Ask them to explain what each question means as they answer. The first useful review happens before launch, when you can still fix a misleading instruction without damaging the comparison.

Check the journey arithmetic before it reaches the calculator

The small wording choices in that questionnaire prevent large errors later. Consider a fictional attendee who travels 4 kilometres by bus, 20 by train and 1 on foot to reach the venue. That is 25 kilometres outward. If the completed return is identical, their total is 50 kilometres across the two directions.

Keep the transport parts separate. The return-inclusive distances are 8 kilometres by bus, 40 by train and 2 on foot. Entering 50 kilometres beside both bus and train would overstate the journey. Entering 25 as though it already included the return would leave half the travel out. Neither error becomes obvious once somebody sees only a final carbon total.

Now take a fictional private-car journey of 100 kilometres in total, with two occupants throughout. If the selected calculation uses a vehicle-distance factor and allocates that vehicle's emissions equally between its occupants, this attendee receives half the vehicle's calculated emissions. The distance allocated to them is equivalent to 50 vehicle kilometres under that method.

If both occupants attend and each answers, those two allocations add back to the same vehicle journey. Don't add a third record for the car on top. If the factor already describes emissions per passenger kilometre with occupancy built in, dividing again would apply an occupancy adjustment twice. Read the factor definition before deciding which arithmetic belongs in your sheet.

Drop-offs need more care. A driver may take someone to the venue, drive home, return to collect them and drive home again. The attendee's occupied distance doesn't capture all that vehicle movement. Ask enough to identify the arrangement, then apply and disclose a consistent allocation rule. Don't pretend that dividing the occupied trip by two resolves every lift.

Work through a changed return before launching the form

Here's another fictional journey to test the questions. An attendee travels from work by train and walks to the venue. After the show, they share a taxi to a friend's home. Their outward train distance tells you nothing about the taxi distance, and the friend's home is a different destination from the office where the trip began.

On the form, they select the train and walking parts outward, then choose that their return was different. They describe the taxi journey and its occupancy separately. The analyst applies the relevant method to each part. If the return field simply repeated the outward trip automatically, the record would describe a train journey that never happened.

Now suppose the attendee planned to take the train home but missed it because the programme finished late. That explanation belongs in the optional difficulty question. It gives your team a cause to investigate, including whether the published finish and travel instructions were realistic. Keep the completed taxi journey in the calculation while preserving the planned train journey as background information.

You can check the timetable and running order to see whether the explanation is consistent with the event's records. You still cannot calculate the emissions saved by an earlier finish from that one response alone. You would need to estimate the alternative journey, consider who could have used it and account for other changes. The useful immediate action is to review the connection before selling the next show.

This is why I would test a changed return before testing a more elaborate question about environmental attitudes. It exposes whether your form can record the experience that matters. If a test respondent cannot explain their actual trip without choosing an inaccurate answer, fix the form before worrying about how polished the report will look.

Reach the audience beyond the easiest people to contact

A ticket-buyer email list is a useful invitation route. It may still miss guests whose tickets were bought by somebody else. Ask purchasers to share the survey with the people who attended, and explain that each person should answer for themselves. Record that distribution method because you won't necessarily know how many individual guests received the forwarded invitation.

You can use Loopyah's attendee email tools to build a message with a link to your survey and send yourself a test before sending it to an appropriate audience. The travel form and analysis still need their own design. Check the applicable sending rules and chosen recipients; a research request isn't permission to add a promotion to every message.

Offer another route for people who don't use the email link. A clearly described web address, a QR code accompanied by readable instructions, or assisted completion can help. Test the form on a phone and with the accessibility needs you expect. Keep assistance neutral so a staff member doesn't steer someone towards the answer that sounds more environmentally responsible.

Avoid recruiting only beside the shuttle stop or cycle parking. Those are convenient places to find people, but they concentrate particular journeys. If staff collect responses in person, plan the locations and times deliberately and record what they covered. A queue for the final train and an early afternoon entrance may contain different audiences.

The American Association for Public Opinion Research's survey research guidance recommends clear, neutral questions and careful attention to who a survey can reach. Apply that to your invitation too. "Tell us how you travelled" leaves room for every answer. "Show us how green our audience is" gives people a result to help you produce.

A reminder may add useful replies, but stop according to the collection plan rather than an attractive percentage. If you offer a participation reward, make it independent of transport type or answer. Record the reward with the method. You need drivers' honest answers as much as cyclists', including answers that make a previous initiative look less successful.

Collect location detail proportionately

Travel information can become personal information when linked with a ticket, email address or a sufficiently distinctive journey. Decide what you actually need before adding identifying fields. A town, broad area or named station may answer the planning question without a full home address or an individual map trace.

Explain who operates the form, the purpose, who can access the answers and how long records will be kept. Make those statements match the actual survey service and your working process. If unique links or response codes connect a reply to a booking, don't describe the survey as anonymous merely because the visible form doesn't ask for a name.

The UK's Information Commissioner's Office explains data minimisation as holding enough relevant personal data for the stated purpose while avoiding unnecessary collection. Its guidance is under review following UK legislative changes. Apply the relevant local rules to your event and survey provider rather than treating a UK page as a universal compliance checklist.

Keep access requests separate from general travel analysis. You can ask what made the journey difficult without asking for a diagnosis. If somebody voluntarily shares sensitive details in a comment, restrict handling and decide what belongs in the analysis. A sponsor reviewing aggregate travel patterns rarely needs the original personal account.

If you need contact details for a prize draw or optional follow-up, explain that purpose and consider collecting them separately from journey answers. Keep only the connection necessary for the process you've described. Plan deletion or anonymisation deliberately, while retaining enough nonpersonal method information to explain the report later.

Clean responses without making them agree with your expectations

Keep the original export and work on a separate checked version. Record the rules you apply to duplicates, incomplete journeys and obvious unit mistakes. If you correct a miles-versus-kilometres entry after confirmation, keep a note of what changed and why. Quietly editing inconvenient answers makes the result impossible to audit.

Useful working fields include a response reference, event date, eligibility, outward and return journey parts, transport type, distance and unit, occupancy where needed, and the source of any estimated distance. Add the calculation status and reason when a response cannot support a complete emissions estimate. The record should explain its own limits.

Don't discard every partial response. Someone may report travelling by train while leaving distance blank. That answer can still support a count of reported train use, although it may not support the carbon calculation without an explicit estimate. Use a different answer count for each result and make the difference visible.

Don't identify duplicates solely because two people report the same journey. Friends in one car should have similar records in an individual-attendee survey. Use the process you designed, such as an explained response reference or a correction route. If an open anonymous form prevents reliable duplicate checks, disclose that limitation instead of inventing certainty.

Check unusual journeys rather than automatically deleting them. A long-distance attendee may be real and may materially change the estimated average. Equally, an extra zero or a distance entered in the wrong unit can distort the result. Seek clarification where permitted, or retain a flagged record and test how it affects the estimate.

Keep operator transport records distinct. Shuttle boarding counts describe passenger trips, potentially with the same person boarding twice. Supplier logs describe vehicle operation, including empty runs. Reconcile them with survey answers under the chosen method. Adding complete shuttle vehicle emissions and a separate passenger-factor estimate for the same service would count that activity twice.

What an incomplete survey can actually tell you

Let's work through a fictional single-evening event, Northbank Sessions. All figures in this example are invented to explain the reporting decisions. They are not Loopyah customer results, expected response rates or travel benchmarks. The event has 800 verified individual attendees, including complimentary guests, and the survey asks each person about their own travel.

After eligibility and duplicate checks, the team has 280 individual responses. Of these, 240 contain sufficient completed journey information for the selected carbon method. The other 40 include useful answers but lack details required for a complete estimate. That means 35% of attendees submitted a response, while 30% supplied a calculable complete journey.

Those percentages use attendance as their denominator. They don't establish an email response rate because the team also distributed an open link through other routes. The number of people who received the invitation is unknown. Keep the description precise enough that nobody mistakes a measure of audience coverage for a measure of successful email delivery.

Suppose the checked calculation for those 240 complete journeys totals 1,440 kilograms of carbon dioxide equivalent. Carbon dioxide equivalent, written CO2e, combines the warming effects of different greenhouse gases into a common measure. Here, 1,440 kilograms is an invented calculation output for the example, not a result derived from published transport factors.

The mean among those complete records is 6 kilograms per person. Multiplying 6 by all 800 attendees gives 4,800 kilograms. The arithmetic is simple. The difficult step is assuming that the 560 people without complete records travelled, on average, like the 240 people whose journeys could be calculated. The survey total alone doesn't justify that assumption.

You can report the estimated 1,440 kilograms associated with the 240 complete respondent journeys, alongside the method and exclusions. You can also show an audience estimate if you have a defensible estimation approach and disclose it. You cannot call 4,800 kilograms the directly measured footprint of all attendees merely because a spreadsheet multiplied it correctly.

Examine who is missing before scaling the result

Compare the respondent group with audience characteristics you can legitimately establish. Event day, broad origin area or ticket group may reveal a mismatch. Use those comparisons where they relate to travel. A difference in premium-ticket representation matters only if there's a reason it could affect the journeys you're estimating.

Perhaps your responses mainly describe local evening visitors, while a substantial part of attendance came for a full weekend. Perhaps the invitation missed guests on group bookings. Identify the actual gap and its possible consequence. Don't assume the missing people drove farther, or that they travelled more sustainably. Their travel remains unknown until evidence supports an estimate.

The House of Commons Library's survey sampling guidance explains why voluntary convenience samples can be unrepresentative. For your event, a larger reply total doesn't remove the selection problem if the same easy-to-reach people keep answering. There's no universal response percentage that makes an opt-in travel survey representative.

Our guide to event survey sample size goes deeper into that decision. Avoid attaching an ordinary random-sample margin of error to an open survey. It won't account for missing groups, unclear distances or the choice people made about whether to reply. A precise-looking interval can leave the biggest uncertainty untouched.

If a sponsor or major spending decision needs a whole-audience estimate, involve someone qualified to assess the survey and carbon method. They may recommend additional collection or separate estimates for relevant groups. They may conclude that the current data support only respondent findings. Find that out before promising a confident audience-wide reduction.

Show what changes when the missing journeys differ

You can explore uncertainty without pretending to know its probability. Continue with the fictional Northbank Sessions numbers. Keep the 240 complete journeys at their estimated total of 1,440 kilograms. Then ask what happens under different assumed averages for the 560 people without complete records.

If those missing journeys average 4 kilograms per person, they add 2,240 kilograms, producing a total of 3,680. At 6 kilograms they add 3,360, producing 4,800. At 9 kilograms they add 5,040, producing 6,480. These are invented assumption tests, not alternative measured results or an established minimum and maximum.

In a real analysis, choose assumptions using relevant evidence and explain why they are plausible. Don't pick a convenient range to make an initiative look successful. An unexplained low estimate and high estimate are still guesses. Keep them separate from any statistical uncertainty assessed under a proper survey model.

Notice what this test can do for an operator. If a costly venue decision looks attractive only under the lowest missing-journey assumption, you need better information before committing. If clearer rail instructions make sense across every reasonable scenario and cost little, you can improve them while acknowledging that the full footprint remains uncertain.

Weighting gives some responses more influence to match known audience characteristics. It may help under a defensible method, but it cannot manufacture journeys for a group that never answered. Keep the original results and document the assumptions. Multiplying a small group's answers until the attendance percentages match doesn't prove that its respondents resemble everybody else in that group.

Choose carbon factors after checking their units and scope

An emissions factor converts an activity, such as a distance travelled, into an estimated quantity of emissions. You need a factor appropriate to the transport, geography, reporting year and calculation boundary. A value copied from another event may use different occupancy assumptions or include different parts of the journey's impact.

The UK government publishes 2026 greenhouse gas conversion factors with guidance and methodology for UK activities. For an event elsewhere, choose an appropriate source for that location or explain a justified fallback. Keep the version, exact factor reference, unit and relevant notes with the calculation so someone else can reproduce it.

Check whether each transport factor uses vehicle kilometres or passenger kilometres. Check what it already includes before adding fuel-supply emissions or another adjustment. Air-travel calculations need their own documented choices, including the treatment of effects beyond the greenhouse gases in the selected factor. Ask the analyst to explain those choices plainly in the method note.

Keep walking, wheeling and cycling visible in the travel results even if the chosen operational method assigns no direct travel emissions to them. Describe that boundary. A zero entry in a particular calculation doesn't establish that the journey, equipment or entire experience had no environmental impact across its life.

Our guide to measuring event environmental impact covers the wider footprint. Attendee travel is one part. Don't let a carefully estimated travel figure become a claim about venue energy, catering or production that the survey never measured. Report the activity and boundary actually covered.

A report paragraph your team and sponsors can judge

Write the finding with its limitations attached. For the fictional example, an internal paragraph could read: "Northbank Sessions had 800 individual attendees. We received 280 eligible individual travel responses, including 240 complete journeys suitable for our calculation. Those 240 journeys produced an estimated 1,440 kilograms of CO2e under the documented method. Participation was voluntary, and we have not established that respondents represent all attendees."

If you share an audience projection, name it separately. "Applying the complete respondents' average of 6 kilograms to all 800 attendees gives a modelled total of 4,800 kilograms. This assumes the 560 attendees without complete calculable records had the same average travel emissions. We have not verified that assumption." Add your real method and uncertainty assessment rather than copying these fictional numbers.

Keep the collection dates, invitation routes, questionnaire version, cleaning rules and factor sources available with that paragraph. Include how return trips, shared vehicles and repeated attendance were handled. State whether supplier records replaced any survey estimates. Someone reviewing the report should be able to follow the method without needing access to personal journey details.

The UK Competition and Markets Authority's Green Claims Code checklist calls for clear, substantiated claims and fair comparisons. Our guide to event sustainability communication helps translate that into attendee and sponsor wording. Keep a claim about one measured activity narrow enough that readers won't interpret it as proof the whole event is sustainable.

A rise in shuttle passengers doesn't by itself prove a carbon saving. You need to understand the journeys replaced, added vehicle operation and other changes. A lower survey average may reflect a more local respondent group. Report what changed in service use while checking whether the evidence supports any stronger explanation.

Give one reviewer the records behind the headline

Before sharing the report, ask somebody who didn't prepare the calculation to reconstruct a small selection of journeys. Include a shared car, several transport types and a changed return. Give them the questionnaire wording, cleaned records and calculation notes. If they need an explanation you haven't written down, add it to the method record.

Have them trace the audience totals too. They should be able to explain why the fictional example contains 280 responses but only 240 complete calculable journeys, and why 560 attendees lack a complete calculable record. Those 560 include the 40 partial respondents as well as the 520 people who supplied no eligible response. Leaving the partial respondents out of both groups would undercount the audience.

Check whether someone reading only the sponsor paragraph could mistake an assumption for an observation. Pay particular attention to labels such as measured, verified, total and saved. The journeys are self-reported, the carbon values are calculated estimates, and a whole-audience projection adds another layer of assumptions. Use wording that preserves those differences when a paragraph is copied into a presentation.

Keep a dated, approved report version with its calculation files. If a late correction materially changes the result, update the calculation and the people who received the earlier figure. Retain a brief explanation of the correction. For a recurring event, that record also prevents next year's team comparing a corrected result with an old number copied from a sponsor deck.

Use the findings before the next booking decision

Explain the reported journey problem to the operations team. If attendees describe a missed connection, check the actual timetable and your finishing time. If people couldn't find the pickup point, inspect the instructions and location. These are decisions you can investigate without claiming to know the travel of every nonrespondent.

For repeat events, retain the core questions and document changes in audience, programme, venue and collection method. Compare totals and clearly defined per-attendee measures, with response coverage beside both. If a better method changes the result, explain that improvement in measurement rather than crediting the entire difference to an operational initiative.

Choose the next missing record according to the decision it could change. That might mean reaching group-booking guests, recording actual shuttle operation or improving the return-journey question. Put an owner and deadline against it before the next survey launches. Your next report should be able to explain what improved in the data as clearly as what changed in the travel arrangements.