Why a Phone Call Beats a Reminder — and Why a Reminder Alone Won't Fix Your Show-Up Rate
A generic reminder call moved nothing in the largest field experiment ever run on this question (N=287,228). A call that got people to state a specific plan moved attendance by 4.1 percentage points among those reached. The difference is the micro-commitment, not the channel.
A reminder notifies. It does not ask anything of the person receiving it. A call built around a micro-commitment asks the registrant to state, in their own words, when they will join and what they will be doing right before it starts. In the largest field experiment ever run on this exact distinction — N=287,228 — the reminder call moved nothing. The micro-commitment call moved attendance by 4.1 percentage points among the people actually reached on the phone. That difference — what the touch asks, not the channel it travels on — is what this post is about.
To be precise about scope: this is not an argument against SMS, and it is not an argument that reminders are worthless. Two well-timed reminders reliably beat one. It is an argument that reminding — however many times, on whatever channel — is a different intervention from asking someone to state a plan, and only the second one has a randomised effect size behind it.
This is not an argument against texting
Do webinar reminders actually raise your show-up rate?
Yes, a little, and it plateaus fast. A three-arm randomised trial of 54,066 patients found that two automated reminders — one three days out, one the day before — produced a 4.4% no-show rate[1], against 5.3% for the one-day reminder alone and 5.8% for the three-day reminder alone (P<.001). Two beats one. Nothing in the published record shows four beating two, and every reminder trial compares a reminder against no reminder at all — while the real comparison for most operators is the email sequence they already run.
Channel does not appear to be the lever either. Swapping the wire a reminder travels on — email, SMS, an automated call — does not appear to move the number on its own. Something else does, and the next section is the experiment that isolated it.
What did the largest field experiment on this ever find?
Nickerson and Rogers[2], Psychological Science, 2010, ran a randomised field experiment across 287,228 people ahead of the 2008 US election. One phone script encouraged the person to vote — a standard reminder call. The other asked them to state a specific plan: when they would vote, what they would be doing beforehand, where they would be coming from. Both were calls. Only one of them asked anything.
The reminder call had no significant effect on turnout. The micro-commitment call produced a +4.1 percentage point increase in turnout among the people actually reached on the phone. The effect across everyone dialed, including the no-answers, was +0.9 percentage points. That gap — same medium, same population, different question — is the cleanest evidence available that asking for a plan is a different intervention from reminding someone of an event, not a stronger version of the same one.
Read the caveat every time this number appears
What is a micro-commitment, and how is it different from a reminder?
A micro-commitment is the registrant stating, in their own words, when they will join and what they will be doing right before the session starts. Not a dropdown they picked, not a restated intention to “definitely be there,” and not a friendly conversation about their business — a specific, self-authored plan.
| Touch | What it asks the registrant to do | What moved in the field experiment |
|---|---|---|
| A one-way reminder (email or text blast) | Nothing — it notifies | Not part of this experiment; see the reminder-ladder evidence above |
| A generic reminder call | Restate that the event is happening | No significant effect |
| The micro-commitment call | State when they will join and what they will be doing right before | +4.1pp among registrants actually reached (ITT +0.9pp) |
The middle row is what most operators already run, dressed up as a phone call instead of an email. The bottom row is the one with a measured effect behind it, and the difference between the two is entirely in what the call asks.
Why does stating a plan out loud work better than being reminded?
The leading explanation is self-persuasion. A meta-analysis of 94 studies (n=8,461)[3] found that forming this kind of if-then plan has a medium-to-large effect on whether people actually follow through on a goal (d=0.65). Someone who has said, out loud, when they will join and what they will be doing right before is defending a plan they authored — not complying with an instruction we gave them. The commitment runs to themselves, which is also why the effect does not obviously depend on who asked the question.
Does the call have to be a live human, or can a disclosed AI ask the same questions?
The nearest evidence on live-versus-automated says live wins on a related outcome. A randomised trial of 9,835 patients[4] found a live staff reminder call produced a 13.6% no-show rate, against 17.3% for an automated message and 23.1% for no call at all. Whether a disclosed AI voice call sits nearer the live arm or the automated arm on the micro-commitment mechanism specifically is not something anyone has published on yet.
Ours says it is an assistant. Then it asks why they signed up — and actually listens. That is a settled decision about how we build the call, not a data point about whether disclosure preserves the effect size.
What we don't know yet
Where does a micro-commitment call fit in a funnel that already sends reminders?
As an addition, not a replacement. Keep the reminders — the reminder-ladder evidence above still applies and still helps a little. What reminders don’t produce is a stated plan, and a stated plan needs a real exchange rather than a broadcast.
Say what it is
The call identifies itself as an assistant before it asks anything else.
Ask why they registered
Opens a thread and produces reusable material for the pitch, not just the show-up number.
Ask for the plan, not the intention
When will you join, and what will you be doing right before it starts — a specific plan, not a restated intention to attend.
Refer back to what they already said
The follow-up references the registrant's own words rather than opening cold a second time.
Frequently asked questions
No. Two well-timed reminders reliably beat one — a three-arm trial of 54,066 patients found a 4.4% no-show rate with reminders at both three days and one day out, against 5.3% and 5.8% for either alone (P<.001). That is real and worth doing. It is also close to the ceiling of what reminding, on its own, can do. The micro-commitment call is a second, separate lever, not a replacement for the first.
A micro-commitment is the registrant stating, in their own words, when they will join and what they will be doing right before the session starts — a specific plan, not a restated intention to attend and not a pleasant conversation about their business. Warmth is worth having, but it is a different thing from a stated plan, and only the stated plan has a randomised effect size behind it.
Treat it as evidence for a mechanism, not a forecast. It comes from a 2008 US voter-turnout experiment, N=287,228, and it is the effect among registrants actually reached on the phone — across everyone dialed, including no-answers, the effect was +0.9 percentage points. Nobody has published a webinar-specific replication, and we don't publish a lift figure of our own for the same reason: we have zero closed clients and no measured effect to show you. A split test on your own list, one cycle, both halves counted the same way, is the honest way to find out.
The nearest evidence says a live call beats an automated one on a related outcome — a trial of 9,835 patients found a live staff call produced a 13.6% no-show rate against 17.3% automated and 23.1% with no call. Whether a disclosed AI voice call sits closer to the live arm or the automated arm on the micro-commitment mechanism specifically is not something anyone has published on.
Maybe — nobody has run that exact comparison. What we can say is that channel, on its own, doesn't appear to be the active ingredient: a Cochrane review of eight trials found SMS and phone-call reminders statistically indistinguishable (RR 0.99, N=2,509), and a separate meta-analysis found two-way texting no better than standard care on attendance (RR 1.03, five trials, 4,374 participants). The evidence points at the ask, not the wire it travels over. We use a call for this specific question because a live conversation lets us hear the answer and refer back to it in the same breath, not because a call inherently reminds better than a text.
The leading explanation is self-persuasion. A meta-analysis of 94 studies (n=8,461) found that forming an if-then plan has a medium-to-large effect on whether people follow through on a goal (d=0.65). Someone who has said out loud when they'll join and what they'll be doing right before is defending a plan they authored, not complying with an instruction.
As an addition, not a replacement. Keep the reminders you already send. What they don't produce is a stated plan, and that needs a real exchange rather than a broadcast — one short call, close to the session, that asks for the plan and then listens.
If reminders are already running and the show-up rate still isn’t where you want it, the next lever with an actual effect size behind it is the stated plan, not another rung on the reminder ladder. Book a call to talk through where a pre-webinar micro-commitment call would sit in your specific funnel. For the stage-by-stage breakdown of what changes and what stays exactly as it is, see what changes for coaches running a live training funnel, and for the rest of the reminder-ladder evidence, how to increase webinar attendance covers the rest of it in more depth.
Founder & Operator, CallHush
Founder and operator of CallHush. The offer is one sentence: you run a webinar, and we increase your show-up rate and your post-webinar sales with an AI voice and SMS system. CallHush has no closed clients yet — the first engagement is a pilot run as a split of the client’s own registrant list, and nothing on this site is presented as a client result.
View all articlesReady to install AI voice on your CRM?
Tell us about your call volume, current CRM, and what you want the AI to handle. If we are a fit we will send a Cal.com link inside 24 hours.