ShowSDROriginal data · two shows · two apps
Conference outreach benchmarks 2026
Between 13 August and 14 September 2026 ShowSDR sent 952 first messages to conference attendees inside two event apps and logged every reply. Pre show messages replied at 26% and produced all 15 agreed meetings. Messages sent during the show replied at the same 26% and produced none. Custom replies kept a conversation alive twice as often as a template (52% vs 24%). One app silently dropped 20% of messages on the two heaviest send days. Below is every number, how it was counted, and what it cannot tell you.
Published 20 September 2026. Data: 1,787 messages across 1,020 threads, two shows, two apps. Download the tables as CSV.
How many attendees reply to a cold message in a conference app?
| Show and app | When first messaged | People messaged | Replied | Asked for a meeting | Meeting agreed by both sides |
|---|---|---|---|---|---|
| Contact.io 2026, Whova | Before the show (Aug 13 to 22) | 543 | 139 (26%) | 68 (13% of messaged) | 15 (2.8% of messaged) |
| Contact.io 2026, Whova | During the show (Aug 23 to 25) | 54 | 14 (26%) | 4 | 0 |
| Affiliate World Americas 2026, iStack | After the show (Sep 10 to 14), connection requests | 351 | 36 accepted (10%), 21 replied | 5 | 1 |
The pre show window is where meetings come from. Attendees use the app to plan their schedules in the weeks before the show; during the show they still reply, but there is no time left to agree a slot. On iStack the first contact is a connection request that must be accepted before a chat exists, so acceptance, not reply, is the first gate, and a request sent after the show cleared it 10% of the time.
Does the wording of the first message matter?
| Contact.io opener (5 variants, ~110 to 121 sends each) | Reply rate |
|---|---|
| Quick one before Denver | 28% (34/121) |
| We run pay per call | 26% (29/111) |
| [Rep] from [Company], pay per call. | 26% (29/112) |
| Straight to it | 24% (28/116) |
| this is [Rep] at [Company], pay per call side | 24% (26/110) |
| Contact.io ask (5 variants) | Reply rate |
|---|---|
| happy to grab 15 minutes in Denver | 31% (36/115) |
| I would take 15 minutes Monday or Tuesday if you have a gap | 30% (36/121) |
| any chance you have 15 minutes at the show | 26% (27/103) |
| worth 15 minutes Monday or Tuesday? | 21% (21/99) |
| free for 15 minutes Monday or Tuesday? | 20% (25/123) |
Openers were a wash. The ask was not: soft statements ("happy to grab 15 minutes") beat direct questions ("free for 15 minutes?") by about 10 points. 28 messages personalised by account replied at 25% (7/28), no better than the template, on a small sample.
Does a custom reply beat a canned one?
| Show | What we sent | They wrote again |
|---|---|---|
| Contact.io | Custom reply drafted for that person | 52% (116/223) |
| Contact.io | Cold template | 24% (139/587) |
| Contact.io | Template for an existing relationship | 29% (4/14) |
| Affiliate World | The client wrote it personally | 54% (19/35) |
| Affiliate World | Two proposed slots plus a booking link | 36% (4/11) |
| Affiliate World | Canned thank you line after acceptance | 12% (2/16) |
| Affiliate World | Day 3 nudge | 0% (0/4) |
Nothing generic came close to a message written for the person. At Affiliate World, after 12 attendees asked a question, the four who got a canned line or nothing went quiet; the ones answered personally mostly continued. The sample is small; the direction is not.
How fast do you have to reply?
| Time from their message to our reply (Contact.io) | Conversation continued |
|---|---|
| 1 to 6 hours | 67% (6/9) |
| 6 to 24 hours | 33% (3/9) |
| 1 to 3 days | 55% (11/20) |
| Over 3 days | 39% (7/18) |
| Never answered | 33% (5/15) wrote again anyway |
Attendees are slow too: from cold message to first reply the median was 2.9 days, 24% within a day, 48% after three days (n = 128). A 24 hour read of any test on a conference app is premature; wait a week.
When should you send?
Contact.io, reply rate by send hour (Pacific): 07h 15% (8/55), 09h 28%, 11h 27% (12/44), 13h 26% (12/46), 14h 34% (18/53), 16h 30% (15/50), 17h 31%, 19h 31% (17/55), 20h 28% (13/46), 22h 37% (11/30). No hour effect; evenings did as well as anything.
Contact.io, reply rate by send day: Aug 13 37% (26/71), Aug 14 26% (38/149), Aug 15 23% (32/137), Aug 16 (Sunday) 4% (1/24), Aug 17 28% (17/60), Aug 18 25% (21/83), Aug 24 36% (9/25), Aug 25 17% (5/29). The two worst weekdays were the two heaviest, and the reason is in the next section.
Affiliate World, acceptance by request hour (Pacific), both batches: 10:00 to 14:00 sends accepted 23% (17/75) in the first batch and 9% (9/103) in the second; 17:00 to 20:00 sends 7% (4/60) and 5% (5/111). Roughly double for late morning and midday, on a global audience, with day and pace changing at the same time. A hypothesis, not a rule.
Is there a volume limit, and what does it look like?
Whova never showed a rate limit message. It did something quieter: on the two days we sent 149 and 137 messages, 97 of them were marked "Failed to send, touch to retry" under our own bubble, visible only inside the thread. Across the campaign 117 of 598 people (20%) never received the message the ledger said was sent. Their reply rate was 6% (7/117) against 30% (146/481) for everyone else. Working assumption: Whova starts dropping direct messages somewhere between 80 and 140 per account per day, and the only way to know is to read the thread after every send. iStack showed no refusal at 38 requests per hour and 214 per day; 24 hour acceptance fell from 11% to 7% when the pace tripled, but the day and the hook changed at the same time.
How fast does the attendee list grow?
The Contact.io directory went from 371 rows on 13 August to 818 on 24 August, the last day of the show. 348 of the 710 people in the final list were not in the first one; 80 of them were never contacted because the list was captured too early. Capture the directory daily until the app closes. Person identity should be the profile id, not name plus company; the company field changed between list and profile 44,724 times in one ledger.
Method and limits
- Every count is a person or a message reconciled against the app's own thread after the campaign, not the send log. "Sent" in a send log means typed on screen; it does not mean delivered.
- Two shows, two apps, two clients, one operator. Verticals: pay per call and performance marketing. Your audience will differ.
- Hook, send hour, pace and calendar day were not independently varied. Where that confounds a comparison, the text says so.
- Reply latency on Whova is an upper bound (first seen times). iStack threads carry no timestamps; only 10 latency pairs exist.
- Meetings held after the show were not systematically captured; 3 are documented, so "agreed" is the last reliable stage.
Questions
What reply rate should an exhibitor expect from cold messages in a conference app?
In our 2026 data, 26% of attendees replied to a pre show cold message on Whova (139 of 543). On iStack, where a connection request must be accepted first, 10% of post show requests were accepted. Expect the pre show number to be the ceiling and the post show number the floor.
Is it better to message attendees before or during the conference?
Before. Pre show messages produced 68 meeting requests and 15 agreed meetings from 543 people; the 54 people first messaged during the show produced 4 requests and no agreed meetings, at the same 26% reply rate.
Do conference apps limit how many messages you can send?
Whova showed no limit message but silently failed 20% of sends on the two days above 130 messages. Read the thread after every send. iStack showed no refusal at 214 requests in a day.
How quickly should I reply to an attendee?
Replies within 6 hours kept the conversation going 67% of the time; after 24 hours the rate fell to 33 to 55%. The median reply in these campaigns took 48 hours, which is the problem ShowSDR exists to fix.
Source: ShowSDR campaign ledgers, Contact.io 2026 (Whova, Denver, August 23 to 25) and Affiliate World Americas 2026 (iStack, Cancún, September 7 to 8), reconciled against the apps' own thread history after each campaign. Full tables and method: Conference outreach benchmarks 2026. Last reviewed 20 September 2026.