The funnel: from invited to attended
Every attendance report starts with the same sequence of numbers, each a subset of the one before. Keep them in this order and give each a precise definition, so that the report reads the same way every time.
- Invited: the people who were actually sent an invitation, not the people on the first list.
- Delivered: invitations that reached an inbox; bounces are listed separately.
- Replied: anyone who answered, yes or no.
- Accepted: those who said they would come, plus the plus-ones they declared.
- Attended: those admitted at the door — scanned or checked in by name.
- No-shows: accepted but not admitted.
- Walk-ins: admitted without an acceptance — no reply, a declined reply, or not invited at all.
Calculating attendance rate and no-show rate
Two ratios do most of the work. The attendance rate is attended divided by accepted; the no-show rate is no-shows divided by accepted. Both use accepted as the base, because that is the number you planned around. A third, the reply rate, is replied divided by delivered, and tells you whether the invitation itself worked.
State the base every time you quote a percentage. “Seventy per cent attendance” means very different things measured against those invited, those who accepted or those who registered, and most arguments about event numbers are really arguments about the base. Count plus-ones consistently, too: either as guests in every figure, or nowhere.
RSVP versus actual attendance
The most useful line in the report is the gap between those who said yes and those who came. It is only measurable if the reply and the door record sit on the same guest record; if replies live in one spreadsheet and the door list on paper, the gap becomes a guess.
Report the gap by name, not only as a number. Two lists come out of it: those who promised and did not come, who are the right audience for a kind “we missed you” note, and those who came without replying, who should be added properly to next year’s list. Over several events the gap becomes a planning figure in its own right: it tells the caterer and the venue how many of your yeses to expect.
Walk-ins, re-entries and exceptions
Keep walk-ins out of the attended total, or show them as their own line. A report that silently adds unregistered guests to the attended figure looks better and is less useful, because it hides how many people your list did not know about. Note who approved each walk-in and on what basis.
Count people, not scans. A guest who leaves for a call and comes back is one attendee with a re-entry, not two attendees. Repeated scans of the same code, manual admissions and overrides belong in a separate exceptions section, with the name of the staff member who approved them. That section is what makes the main numbers credible.
Breaking the report down by group
A single total hides most of what the report should say. Split every figure by the segments you used to plan: VIP, press, partners, employees, speakers, first-time guests, plus-ones. A strong overall turnout can hide a press table that stayed empty or a partner group that declined in a block.
Add the dimensions that matter for your format. For a conference, attendance per session or per day; for an event with several entrances, arrivals per door and per hour, which tells the team where to place staff next time; for a public programme, the share of new guests against regulars.
- By segment: VIP, press, partners, employees, speakers, plus-ones.
- By time: arrivals per quarter-hour, the peak, late arrivals.
- By door or zone: where queues formed.
- By session or day, for multi-track events.
What to report to whom
Different readers need different reports from the same data. Write each one for its reader, and never change the underlying numbers between versions.
- Leadership: one page — attended against target, the RSVP gap, one or two lessons, the decision you recommend.
- Operations and venue: arrivals by hour and door, peak queues, catering numbers against actual, walk-ins and exceptions.
- Sales or account teams: the names of key guests who came and those who did not, by account.
- Sponsors: totals they were promised — attendance, visits to their stand, leads gathered with consent — without guest-level personal data unless guests agreed to share it.
- The client, for agencies: the full funnel, the gap and the segment view, in the client’s own branding where possible.
Exporting and keeping the report
Produce the report the next day, while questions can still be answered from memory. Export it in the form the reader accepts: a designed PDF for leadership and clients, a spreadsheet for operations and anyone who wants to filter. Keep the definitions on the page so that the numbers mean the same thing a year from now.
An attendance report contains personal data about every guest. Share named lists only with people who need them, prefer totals wherever a total answers the question, and keep the report with the event so next year’s planning starts from facts rather than from recollection.
Where the numbers come from
The report is only as good as the record behind it. Replies collected through personal links, check-ins scanned at the door and walk-ins added on the spot all need to land on the same guest record. In Winglark, the reporting page draws on exactly that — replies against who was admitted, by name and by group — and exports the report as a PDF, a CSV file or an Excel sheet.