Invitation and RSVP benchmarks, from real data
Most "invitation trends" roundups are guesswork. This one is measured. Every finding below was recomputed from production on September 10, 2026, across thousands of events sent through InviteDrop, and every rate counts the same population: guests who were individually sent an invitation through the platform.
That scoping matters more than it sounds, and it is where most published RSVP statistics go wrong. An individually sent guest list is the only complete list: it contains every person the host meant to invite, including the ones who decline and the ones who never answer. Replies also arrive from guests who were never sent to directly, because someone forwarded them a shared link. Those are real guests, but they are self-selected. A person who is not coming rarely bothers to open a link they were forwarded, so their decline never enters the data at all. Include them and the acceptance rate rises by several points without one extra guest saying yes. They are excluded from both sides of every fraction here, and finding 8 shows exactly how large that distortion is.
Quick answer: about half of sent invitations get any reply at all, about 78% of those replies are a yes, delivery is split almost evenly between email and text rather than dominated by email, and replies arrive in the first day or usually not at all.
1. About half of invitations never get a reply
Roughly 51% of sent invitations receive a response. The other half go unanswered, which is the single most useful number on this page and the one most platform marketing quietly skips.
It reframes what an RSVP tool is for. The value is not tallying the people who replied, which a spreadsheet does perfectly well. The value is showing you the half who did not, individually, so you can chase them. A platform that only reports the yes count is hiding the harder half of the job.
2. When guests reply, about 78% say yes
Among guests who responded:
- About 78% accepted.
- About 22% declined.
- Almost nobody chose "maybe", at well under one percent of replies.
A previous version of this post reported 92% acceptance. That figure was wrong, and it was wrong in the direction that flattered us: it blended self-selected shared-link traffic into a rate presented as though it described invited guests. The corrected number is meaningfully lower, and it changes the planning advice. Roughly one reply in five is a no, so a responded headcount is not a final headcount, and the "maybe" option that almost nobody uses is a nicety rather than a feature worth choosing a platform over.
3. About a third of accepting guests bring someone
Of guests who accepted, about 34% added at least one other person, and those who did brought about 1.5 extra people on average. That is fewer guests bringing more people each than we previously reported.
The practical consequence is unchanged and still underrated: a headcount taken from acceptances alone runs meaningfully low. Decide your plus-one policy before you send rather than after the replies start arriving, and set a per-guest limit so the total stays predictable. InviteDrop caps plus-ones per guest and tracks the running total for you.
4. Email and text are almost evenly split
By delivery channel, invitations divide close to three ways:
- About 41% are addressed to an email address only.
- About 42% are addressed to a phone number only.
- About 17% carry both.
This is the finding that changed most, and the one worth acting on. A previous version of this post claimed nine in ten invitations went out by email. That was wrong by a wide margin. Text is not a minority channel or a fallback; it is now the marginally larger one.
The implication when choosing a platform is direct. A service that sends by email alone is simply unable to reach roughly two in five of the guests people actually want to invite, and asking a host to collect email addresses for guests they only have phone numbers for is asking them to do the platform's job. InviteDrop sends by email and by text from the same guest list, and every invitation also has a shareable link that works in any messaging app.
5. Replies arrive on the first day or usually not at all
Among guests who received an invitation and responded:
- About 23% replied within an hour of the invitation being sent.
- About 46% replied within 24 hours.
- About 70% replied within a week.
The median reply takes about 39 hours. Read those together and the shape is clear: nearly half of all the replies you are ever going to get land on day one, and the curve flattens hard after the first week. A guest who has not answered within a day is unlikely to answer at all without a nudge.
So the first day after you send is your highest-response window, and a reminder is not optional housekeeping. It is the mechanism that recovers a real share of the half from finding one. Send when guests are likely to be on their phones, then let a reminder do the chasing.
6. Hosts plan further ahead than digital invitations get credit for
Among hosts who set an event date and actually sent their invitations, the median invitation goes out about 43 days before the event. About 35% send two months or more ahead, and only about 7% send inside the final week. This is the smallest population on this page, since it covers only events where the host filled in a date, so treat it as directional rather than precise.
It still undercuts a common assumption. Digital invitations are often framed as the last-minute option, the thing you reach for when there is no time to print. The measured behaviour is the opposite for most hosts: they are planning over a month out and choosing digital anyway, for the tracking and the reach rather than the speed. The speed matters for the minority who need it, and nothing else can offer it.
7. Most guest lists are small
About 75% of guest lists are 25 people or fewer, and about 96% are 100 or fewer. Large events exist but they are rare.
That is worth knowing before you compare prices, because pricing models diverge sharply at the small end. Per-recipient pricing, including credit or coin systems, is designed around the rare large list and tends to look expensive per head on a common small one. A flat price per event does the reverse. Our own pricing is a one-time Event Pass per event for exactly this reason, and the first five invitations on any event are free.
8. The acceptance rate you see quoted is probably inflated
This one is about how to read everyone else's numbers, including the ones we published before today.
Guests who arrive by following a shared link accept at about 91%. Guests who were sent the invitation directly accept at about 78%. Same platform, same designs, same week: a thirteen point gap produced entirely by which guests end up in the dataset.
The mechanism is selection, not enthusiasm. When a host sends to a list, everyone on that list is measured, including the people who say no and the people who ignore it. When an invitation spreads by a forwarded link, the only people who ever reach the RSVP form are people interested enough to open it. Everyone else silently never appears, so their decline is never counted.
This is worth knowing whenever you see a platform advertise a high acceptance or response rate. Ask which guests were counted. A rate computed over link traffic is a survivorship artefact, and it will always look better than the rate you will personally experience when you send to your actual guest list. The full method, and the rest of the figures, are on our RSVP benchmarks page.
What this means for your next invitation
Put the findings together and they point somewhere more specific than "digital is convenient". Half your guest list will not reply without being asked twice. Two in five of them are reachable by text rather than email. Almost all of the replies you get will arrive in the first day. And a third of the people who say yes are bringing someone you have not counted yet.
Those are the four things a digital invitation should handle for you: reach guests on the channel they actually use, show you who has not replied, chase them automatically, and keep the real headcount rather than the optimistic one. You can create an invitation on InviteDrop and start collecting RSVPs in minutes.
Method: recomputed from production on September 10, 2026, across thousands of events sent through InviteDrop. Rates count only guests who were individually sent an invitation; guests who arrived by a forwarded link are excluded from both numerator and denominator, for the reason given in finding 8. Figures are rounded and will shift as more invitations are sent. This post previously reported an occasion mix and several figures that were not supported by the underlying data; those findings were removed or corrected on September 10, 2026 rather than left standing.
