How to Increase Webinar Attendance: 6 Tactics Ranked by Evidence
Webinar attendance rate - the share of registrants who join the live session - is the number most operators try to fix with reminders. The evidence says reminders are the smallest lever available, and the biggest one is a decision you make before a single registration comes in: whether the event is free. This page ranks six tactics by the strength of the evidence behind each, names the source and the sample for every figure, and labels the ones that are a single operator's observation rather than a study.
A page published by a company that sells AI calling software should probably open by telling you to make more calls. It does not, because that is not what the evidence supports. The largest observed gap in show-up rate in everything we reviewed comes from charging a token fee, and the second largest comes from changing the format of the event. Calls and reminders sit fourth and fifth, and they are worth roughly one percentage point each in the best-controlled study available.
TL;DR
Ranked by evidence, strongest lever first: (1) charge a token registration fee - one organiser's six years of monthly events shows ~30% show-up free versus ~85% paid, which is a single-operator observation and not a study; (2) run the session as a working session, not a lecture, so the replay stops being a substitute; (3) ask for a micro-commitment at registration; (4) send two reminders, not one (Steiner 2018, 3-arm RCT, N=54,066: 4.4% missed versus 5.3% and 5.8%, P below .001); (5) ask for a concrete plan on a call (Nickerson & Rogers 2010, N=287,228: +4.1 points among those actually reached); (6) adding more touches does not work - a call is statistically indistinguishable from a text (Cochrane 2013, RR 0.99) and two-way texting is null for attendance (RR 1.03). All figures below are from US or English-language sources.
Which webinar attendance tactics actually have evidence behind them?
Here is the uncomfortable shape of this topic: the levers with the biggest observed effect have the weakest evidence, and the levers with the strongest evidence have the smallest effect. Anyone who tells you otherwise is selling something. The table below is the whole article in one view.
| # | Tactic | Best available evidence | Evidence strength |
|---|---|---|---|
| 1 | Charge a token registration fee | One organiser, six years of monthly online events: ~30% show-up free vs ~85% paid | Single-operator observation - NOT a study |
| 2 | Run it as a working session, not a lecture | One operator: it "shifted the perceived value from I can catch the recording to I need to be there" | Single-operator observation - NOT a study |
| 3 | Ask for a micro-commitment at registration | Two operators describe the same mechanic independently in one thread | Operator consensus, no controlled test |
| 4 | Send two reminders, not one | Steiner 2018: 4.4% missed (3-day + 1-day) vs 5.3% (1-day) vs 5.8% (3-day), P below .001 | Randomised, N=54,066, US healthcare |
| 5 | Ask for a concrete plan on a call | Nickerson & Rogers 2010: +4.1pp among those contacted; a standard encouragement call had no significant impact | Randomised, N=287,228, 2008 US election |
| 6 | Add more touches and more channels | Cochrane 2013: a call is indistinguishable from a text (RR 0.99). Odegard 2022: two-way SMS null for attendance (RR 1.03) | Evidence points AGAINST it |
Read the strength column before you read the ranking
Items 1-3 are reports from named operators on public forums. They have no sample size, no control group, and no linkable permanent record we could verify, so we present them as observations rather than results. Items 4-6 are randomised controlled trials in healthcare and political science with samples in the tens of thousands. Neither category can be substituted for the other: the strong evidence is about a small effect, and the big effect has weak evidence. Both facts are load-bearing.
1. Does charging a registration fee increase webinar attendance?
The single largest gap in show-up rate we found anywhere is between free and paid registration, and it comes from one event organiser reporting six years of monthly online events. Their numbers: a 70% no-show rate from free registrations and a 15% no-show rate from paid registrations - roughly 30% versus 85% show-up - and, in their words, "consistent across the board as far as topics, time of day, day of the week."
| Registration type | Reported no-show rate | Implied show-up rate |
|---|---|---|
| Free registration | 70% | ~30% |
| Paid registration | 15% | ~85% |
What this is, precisely
This is one organiser's own operational data, self-reported on a public forum, covering six years of monthly events. It is not a study. There is no control, no disclosed sample size, no independent verification, and the permalink is not reliably retrievable, so we quote it without a link. It could be confounded by audience type, topic, or price point - a person who pays to attend is a different person from one who clicks a free registration button. We publish it anyway, and first, because a ~55-point spread from a single decision dwarfs everything that is properly evidenced below, and ignoring it because it is inconvenient to a calling product would be dishonest.
For context on the free side of that number: Goldcast's 2025 B2B benchmark, covering 19,531 webinars from 418 brands and 3.5 million registrants, reports a 33% attendance rate - while also reporting an average of 238 registrants and 51 attendees per webinar, which is 21% (Goldcast, 2025, US/global B2B). Both numbers are in the same report. Roughly a third, at best, is where free registration lands - which is exactly the number the organiser above reported before they started charging. We unpack why published benchmarks disagree so violently in the webinar attendance statistics review.
The practical version is a token fee, not a real price: a small amount that costs the registrant something and costs you almost nothing to collect. It filters the list rather than growing it - you will get fewer registrations - so it is a trade you make deliberately when you would rather have 80 people who show than 400 people who registered.
2. How do you compete with your own replay?
The most-upvoted reason we found for skipping webinars is not forgetfulness. It is this, from a marketing forum, the highest-voted answer in its thread: "I don't go to webinars because everyone sends a recording now. Now I can access the exact info when I need it." No reminder cadence answers that. The registrant is not confused about the time; they have made a rational decision that the live event and the recording are the same product, and one of them is cheaper.
The only tactic in our whole review that addresses it directly came from an operator describing a format change: they shortened the session, made it interactive, and - in their words - started "framing the live event as a working session rather than a lecture, which shifted the perceived value from ‘I can catch the recording’ to ‘I need to be there to get the most out of it.’"
The test to run on your own event
Ask: if someone watches the recording at 2x speed tomorrow, what exactly do they lose? If the honest answer is "nothing", your show-up rate problem is a format problem and no amount of reminding, texting or calling will move it. Things that only exist live: answering their specific situation out loud, feedback on work they do during the session, a decision they leave having made, a limited-availability offer that expires with the session. Things that do not: slides, a framework, a case study, a demo.
This is also a single-operator observation with no sample and no control. We rank it second because it is the only idea in the corpus that attacks the number-one stated reason people skip, and because the mechanic is cheap to test on a single event.
3. What is a micro-commitment at registration?
A micro-commitment - also called a micro-yes or pre-work - is one small action you ask the registrant to take immediately after they register, beyond handing over their email address. Three operators in a single public thread described the same mechanic independently, without prompting, in almost the same words. That convergence is the reason we use their vocabulary rather than the academic term.
The mechanic in operators' own words
Version A: "In the confirmation email, ask them to click one button: ‘I'm coming live’ + pick 1 question they want answered. That micro-yes boosts show-up rates a lot."
Version B: "Add a ‘pre-work’ micro-commitment: right after registration, ask one question they must answer (1-click poll): ‘What's the #1 thing you're hoping to get from this?’ Then open the webinar by referencing the top answers."
Two independent operator descriptions from one thread. No sample size, no control group, no measured effect. Treat as a hypothesis worth testing, not a benchmark.
Two things make this worth the small amount of work it costs. The first is that the answers are useful on their own: a list of what 300 people said they are struggling with, in their own words, is reusable in the session, in the follow-up, and in the next campaign - whatever it does to attendance. The second is that it is the one place in the funnel where you are already asking the registrant to do something, so the marginal friction is near zero.
What we will not claim is a number. Nobody has run a controlled test of a registration micro-commitment on webinar attendance and published it. The nearest properly evidenced relative is the plan-making research in section 5, which is a different intervention delivered on a different channel.
4. How many webinar reminders should you send?
The best-controlled answer to this question comes from healthcare, not marketing, and it is two. Steiner and colleagues ran a three-arm randomised trial across 25 primary-care clinics with 54,066 patients, comparing an automated reminder 3 days before, a reminder 1 day before, and both.
| Reminder arm | Missed-appointment rate | High-risk patients |
|---|---|---|
| 3 days + 1 day before (two reminders) | 4.4% | 20.5% |
| 1 day before only | 5.3% | 24.2% |
| 3 days before only | 5.8% | 25.0% |
All differences were significant at P below .001 (Steiner et al., American Journal of Managed Care, 2018, US primary care). Note what the study actually delivers: two automated reminders beat one, and the absolute gap is about one percentage point. That is a real, replicable, statistically robust effect - and it is small. It is also the single most common thing sold as a webinar attendance solution.
Two caveats before you port it. The population is patients with a booked medical appointment, which is a stronger prior commitment than a free webinar registration. And the base rates are nothing like a webinar's: a 4.4% no-show rate is a world away from the 70% reported above for free registrations. The direction - two beats one, and a single reminder timed close beats one timed far - is what transfers. The magnitude does not.
What this means for a reminder sequence
Send a confirmation at registration, then two reminders: one a few days out and one on the day. If you are currently sending five, the evidence does not support the extra three, and section 6 shows what happens to operators who add them anyway. If you are sending one, add the second - it is the cheapest properly evidenced improvement on this page.
5. Does calling registrants before the webinar increase attendance?
It depends entirely on what the call says. The strongest evidence on this question is not from marketing at all - it is a field experiment on 287,228 people during the 2008 US presidential election. Nickerson and Rogers tested phone scripts that asked people to form a concrete plan: what time they would go, where they would be coming from, what they would be doing beforehand.
The abstract is explicit on both halves of the finding: forming a voting plan "can increase turnout by 4.1 percentage points among those contacted", while "a standard encouragement call and self-prediction have no significant impact" (Nickerson & Rogers, Psychological Science, 2010). The contrast is the whole point. Ringing someone up to say "don't forget, it's tomorrow at 2" is the arm that did nothing. Asking them to say out loud when and how they will actually join is the arm that worked.
Three caveats that must travel with this figure
1. It is the treated-only estimate. The 4.1 points is measured among registrants you actually reach. If you connect with a third of your list, the effect on the whole list is proportionally smaller.
2. It used live human callers. Whether the same effect survives when the caller is a disclosed AI is empirically unestablished - nobody has published on it. We are not going to assert the transfer as proven when it has never been tested.
3. The setting is a US presidential election in 2008, not a webinar. Voting and joining a webinar are both "be somewhere at a time" behaviours, which is why the mechanism plausibly carries. That is an argument, not a measurement.
The operational translation is short: if you are going to call registrants, the call has to ask a question and wait for an answer. A call that only repeats the time is, on this evidence, the arm that had no significant impact. We describe how we structure that conversation on the webinar reminder calls page, and what happens to attendees and no-shows afterwards on post-webinar follow-up calls.
Whether you may call at all is a separate question with a jurisdiction-specific answer. In the EU and UK, a registration form does not by itself buy you the right to phone someone, and the rules on numbering and consent differ by country - we go through that in the EU/UK legality guide and on the registrant consent page. US and other markets run on different rules entirely; check yours before you dial.
6. Does adding more touches and more channels fix a low show-up rate?
No, and this is the tactic we would most like to be true, because it is the easiest thing to sell. The evidence runs against it from two directions: controlled trials, and operators who have already tried it.
| Study | Design and sample | Finding | What it means here |
|---|---|---|---|
| Cochrane 2013 (Gurol-Urganci et al.) | International healthcare. 7 studies, 5,841 participants for the no-reminder comparison; 3 studies, 2,509 for the call comparison (the review has 8 trials, 6,615 participants in total) | SMS vs no reminder RR 1.14 (1.03-1.26). SMS vs a phone call RR 0.99 (0.95-1.02) | A reminder beats nothing. A call does NOT beat a text - and costs more |
| Odegard 2022 (PLoS One) | 5 trials, all set in Sub-Saharan Africa. The review reports both 6,627 and 4,374 participants for this same analysis - the discrepancy is in the source | Two-way text vs standard care, attendance: RR 1.03 (0.95-1.12), I2 53% - null | Conversational texting shows no measurable attendance gain. Healthcare in Sub-Saharan Africa is a long way from a European webinar, so transfer is uncertain in both directions |
| Parikh 2010 (Am J Med) | RCT, N=9,835, US outpatient clinic | No-show 13.6% live staff / 17.3% automated / 23.1% none, P below .01 | A live human call beats an automated one. Nothing here tested a disclosed AI |
| Steiner 2018 (Am J Manag Care) | 3-arm RCT, N=54,066, US primary care | 4.4% (two reminders) vs 5.3% and 5.8% (one) | Two beats one. The study did not test three, four or five |
Two independent operators in our review ran the full stack - email, SMS, and a human phone call - and it did not save them. One described the sequence as: follow-ups by email and text, a day before, several hours before, an hour before, then a phone call five minutes before. The result was 3 of 15 showed up, and zero converted.
Label this case honestly
That 3-of-15 case is booked one-to-one appointments, not a webinar, and the source is a forum post without a retrievable permalink. It is not a benchmark and we do not present it as one. It matters for a narrower reason, and the narrow reason is the strong one: if a human calling five minutes before does not fix a show rate, an AI calling cannot claim to either. Any vendor - us included - who tells you their calls are the difference between a 30% and a 70% show-up rate is describing something nobody has demonstrated.
So the honest position on channel is this: reminders beat no reminders, two beat one, and beyond that the marginal touch buys you very little and costs you goodwill. The thing that changes outcomes is not the number of messages - it is whether any of them asks the registrant a question they have to answer.
What does the whole playbook look like, in order?
Six steps, in the order you would actually execute them, with the evidence label attached to each so you know which ones you are betting on and which ones are safe.
Decide whether the event should be free at all
A token registration fee is the largest observed lever on show-up rate in everything we reviewed - one organiser reported ~30% show-up free versus ~85% paid across six years of monthly events. It is a single-operator observation, not a study, and it will shrink your registration count. Test it on one event before you conclude anything.
Design the session so the replay is not a substitute
The most-upvoted reason people skip webinars is that a recording is coming. Shorten the session, make it interactive, and frame it as a working session rather than a lecture. If nothing is lost by watching it tomorrow at 2x, no reminder will fix the attendance.
Ask for one micro-commitment on the registration page
One click or one typed answer immediately after registration: "I am coming live", or "what is the number one thing you want out of this?". Operators call this a micro-yes or pre-work. No controlled test of it exists for webinars, but the answers are independently useful and the friction is near zero.
Confirm immediately, then send exactly two reminders
A confirmation at registration, one reminder a few days out, one on the day. In the largest randomised test available (Steiner 2018, N=54,066) two automated reminders beat one - 4.4% missed versus 5.3% and 5.8%, P below .001. The study did not test three or more, and operators who send five report no better outcomes.
If you call, ask for a plan - never just repeat the time
Nickerson and Rogers (2010, N=287,228) found a plan-eliciting call raised turnout 4.1 points among people actually reached, while a standard encouragement call had no significant impact. Ask when they will join, from what device, and what they will be doing right before. Check the rules in your market before calling registrants at all.
Measure show-up rate by traffic source, because nobody has published a benchmark for it
There is no published webinar show-up rate benchmark segmented by traffic source. Cold paid traffic and a warm house list are not comparable, and any single industry-average figure quietly averages the two. Your own numbers, split by source, are the only benchmark that means anything for your decisions.
What can you not fix with reminders?
Three things, and they are the three most likely to be your actual problem.
Cold traffic. If your registrants came from a cold ad and have no prior relationship with you, reminders will not rescue the show-up rate. Individual operators report show-up rates in the low single digits on cold lists even after email, SMS and phone - forum reports, not measurements, and we have found no study that puts a number on it. We concede this plainly because the alternative is selling a reminder sequence to someone whose problem is upstream of it - the offer, the audience, or the ad.
The replay. The number-one stated reason people skip is that a recording is coming, not that they forgot. A reminder addresses forgetting. It does not address a rational decision that the recording is the better deal. Only the format change in section 2 touches that, and it is unproven.
And one thing nobody knows, including us
There is no published performance data on AI voice calls as event reminders, with any disclosed methodology. Not from us, not from any vendor we could find - no sample size, no control group, no measurement period, in any case study we reviewed. What is known is adjacent: a live human call beat an automated reminder system in a US randomised trial (13.6% versus 17.3% no-show, Parikh 2010, N=9,835), and a text was statistically indistinguishable from a phone call in the Cochrane pooled comparison. Where a conversational AI call lands between the robocall and the human is an open question that nobody has answered in public. Anyone quoting you a number for it made it up.
We would rather say that than invent a figure. If you want the same argument applied to appointments rather than webinars, we went through that literature in how AI receptionists reduce no-shows - and reached a different conclusion on channel there, leaning on the hospital-reminder reviews that favour the phone, while this page follows the pooled call-versus-text comparison, which is null. Read the two against each other rather than treating either as settled. The reasons registrants ghost are in why webinar registrants don't show up. If you want to talk through what is realistic for your own list before you buy anything, our contact page is the place - and the AI webinar attendance page sets out what our own system does and, more usefully, what it does not.
Frequently Asked Questions
Frequently Asked Questions
Ranked by the strength of the supporting evidence: charge a token registration fee (the largest observed effect, though it rests on one organiser's six years of data rather than a study); design the session so the replay is not a substitute; ask for one micro-commitment at registration; send two reminders rather than one (Steiner 2018, randomised, N=54,066); and if you call, ask registrants for a concrete plan rather than repeating the time (Nickerson & Rogers 2010, N=287,228). Adding more touches and more channels is the one tactic the evidence argues against.
The largest single gap we found comes from one event organiser reporting six years of monthly online events: a 70% no-show rate on free registrations versus 15% on paid ones, which is roughly 30% versus 85% show-up. This is a single-operator observation self-reported on a public forum, not a study - no control group, no disclosed sample, no independent verification - and it may be confounded by audience type or topic. It is still the biggest effect in the corpus, so a token fee is worth testing on one event, accepting that registration volume will fall.
Two. In the largest randomised trial available (Steiner et al., American Journal of Managed Care, 2018, three arms, 54,066 US primary-care patients), a reminder 3 days before plus one 1 day before produced a 4.4% missed-appointment rate, versus 5.3% for the 1-day reminder alone and 5.8% for the 3-day reminder alone, P below .001. Two automated reminders beat one, and the absolute difference is about one percentage point. The trial did not test three or more reminders, and operators who send five report no better results.
The best pooled evidence says a call does not beat a text. The 2013 Cochrane review of mobile phone messaging reminders found, across 7 studies and 5,841 participants, that SMS reminders improved attendance versus no reminder (RR 1.14, 95% CI 1.03-1.26) - and across 3 studies and 2,509 participants, no detectable difference between an SMS reminder and a phone call reminder (RR 0.99, 95% CI 0.95-1.02), while two of the included studies reported the cost per text message per attendance to be 55% and 65% lower respectively than the cost per phone-call reminder. (The review includes 8 trials and 6,615 participants in total; those two comparisons are subsets of it.) A 2022 PLoS One meta-analysis found two-way text messaging produced no improvement in appointment attendance versus standard care (RR 1.03, 95% CI 0.95-1.12), across trials set entirely in Sub-Saharan Africa. All of this is healthcare data, which is the best-studied analogue available - it is not webinar data, and nobody has published the webinar equivalent.
The most-upvoted reason we found in operator and audience discussions is not forgetfulness - it is the replay. People know a recording is coming and treat it as an equivalent product available at a cheaper time cost. Other frequently cited reasons are workload, webinar fatigue, and a rising bar for what is worth 45 minutes. Reminders address forgetting, which means they address a problem that is not the main one.
Only if the call asks a question. Nickerson and Rogers (Psychological Science, 2010) ran a field experiment with 287,228 participants during the 2008 US presidential election and found that helping someone form a concrete plan raised turnout by 4.1 percentage points among those actually contacted, while a standard encouragement call and self-prediction had no significant impact. Three caveats matter: the 4.1 points is the treated-only estimate, so it applies to registrants you actually reach; the callers were live humans; and the setting was an election, not a webinar.
Nobody has published performance data on AI voice calls as event reminders with any disclosed methodology - no sample size, no control group, no measurement period, from any vendor we could find, ourselves included. What is known is adjacent: a live human reminder beat an automated reminder system in a US randomised trial (13.6% versus 17.3% no-show, Parikh 2010, N=9,835), and the plan-making research that motivates the approach used live human callers. Whether the effect survives with a disclosed AI caller is empirically unestablished. Treat any specific number you are quoted for it as unsupported.
Published benchmarks range from about 21% to about 60%, and the spread is mostly explained by who supplies the registrants and how attendance is defined rather than by real performance differences. Goldcast's 2025 B2B report, covering 19,531 webinars, states a 33% attendance rate while also reporting 238 average registrants and 51 average attendees per webinar, which works out to 21%. Critically, there is no published show-rate benchmark segmented by traffic source, so no industry average tells you what a cold-ad audience should do versus a house list.
Founder & CEO, AInora
Building AI digital administrators that replace front-desk overhead for service businesses across Europe. Previously built voice AI systems for dental clinics, hotels, and restaurants.
View all articlesReady to try AI for your business?
Hear how AInora sounds handling a real business call. Try the live voice demo or book a consultation.
Related Articles
Webinar Attendance Statistics: Every Study (2026)
Why published show-up rates range from 21% to 60%, which benchmarks are structurally comparable, and which widely quoted figures are fabricated.
Why Webinar Registrants Do Not Show Up
The ranked reasons registrants ghost - the replay first, not forgetfulness - in the words of the people who skip.
Is It Legal to Call Webinar Registrants? (EU / UK)
What a registration form does and does not authorise under EU and UK rules, country by country.
How AI Receptionists Reduce No-Shows
The appointment-side reading of the same evidence - it leans on the hospital-reminder reviews that favour the phone, where we follow the pooled call-versus-text comparison that is null.