Marketing a Webinar: The Registration Is Not the Result
Written by Martin H.
You filled the room, and half of them showed up. A quarter stayed to the end, and the follow-up went out four days late. Which number did you actually get paid for?
I have sat in more than one webinar post-mortem where the registration number was the only thing anybody could quote. A big room, a number in the deck that made the quarter look handled, and a follow-up review three weeks later where nobody could say what had been sent to the people whose ticket was really for the template rather than the session. The room had been full. The pipeline had not moved. Nobody could point to the decision that separated those two facts, because no decision had been made.
What a Webinar Is Actually For (and What It Cannot Do)
A webinar does one thing better than any other format: it lets a stranger watch you think for forty-five minutes before they have to trust you with anything. That is the whole asset. It is not reach, it is not a list build, and it is certainly not a lead magnet with a calendar invite attached.
The format cannot do three things, and most of the disappointment in this channel comes from asking it to.
It cannot qualify people. Registration is a low-cost action: an email address and a time slot. If you want to know whether someone should be your customer, a webinar is a bad instrument, and treating sign-ups as intent is how a demand team ends up with a list it cannot work. The qualification question belongs to a designed gate, which is the argument in our piece on lead generation funnels, and it does not belong here.
It cannot create demand that does not exist. If nobody is looking for what you sell this quarter, a webinar will gather the small number who are and then stop.
And it cannot survive a promise it did not make. This is the failure I keep meeting, and it is the one the readable advice on this topic leaves out. More on that below, because it is the hinge the rest of this article turns on.
Promotion That Selects for Attendance Quality
Every promotion channel fills a room with a slightly different crowd. That is not a marketing platitude. It is a mechanical consequence of what you offered in exchange for the registration.
Two offers sit at opposite ends. Offer the session. "We will spend forty-five minutes on how to run a webinar that produces pipeline." The person who registers is buying the hour. Offer a thing. "Register and get our 47-page planning template." The person who registers is buying the template, and the template is delivered the moment they hit submit. Their reason for attending has already been satisfied. When the reminder arrives, they have no reason to clear the slot.
The second offer will always produce more registrations. It is also the offer that makes your follow-up unsellable, because the only thing you know about that person is that they wanted a template. You cannot write a useful follow-up to a template collector. You can only write a generic one, which is what most teams end up sending.
Channel | Who it selects for | What you pay per attendee |
|---|---|---|
Your own list, offer = the session | People who already trust the hour to be worth it | Cheapest. No media cost, highest attendance rate, smallest room |
Your own list, offer = a download | Template collectors and list-builders | Cheap per registration, expensive per attendee: the room grows and the rate falls |
A co-host's audience | Borrowed trust that has not been earned yet | Cheap to reach, weak on attendance; you are a stranger's recommendation |
Paid social, cold | Anyone the algorithm can find at the price you set | Most expensive per attendee by a wide margin |
Paid search on competitor and category terms | People actively comparing options | Expensive per registration, unusually strong attendance, because the query was a decision |
The third column is an ordering, not a benchmark. I looked for a published per-channel webinar cost per attendee and did not find one, so nothing in that column is a figure I can defend as a number. Read it as cheapest to most expensive and stop there.
Read the rows in order and the pattern is uncomfortable. The two cheapest ways to grow a registration number are the two that add people with the weakest reason to attend. That is why registration volume feels like progress while the room gets quieter.
The Attendance Drop and How to Read It
The registration-to-attendance rate is the most stable number in this channel and the most ignored. Contrast's 2026 benchmark, drawn from over a million registrants, puts live attendance at 44 to 50 percent. Other 2026 platform reads land close: Livestorm reports 47.7 percent, and a dataset summarised by blog.mean.ceo puts the median at 41.6 percent. Plan against the lower end unless your audience is genuinely warm.
The number that should change what you do sits one layer down. In the same Contrast data, webinars under 100 registrants run about six percent higher on live attendance, and webinars with more than 1,000 registrants drop to 37 percent. The bigger the room you buy, the worse the rate. That is not a quirk of one platform; it is what happens when promotion reaches people whose reason for registering was thinner.
Then the drop inside the hour. The same 2026 summary puts average live viewing duration at 37 minutes. That single figure tells you where your offer has to sit, and it is earlier than most presenters are comfortable with. If your ask arrives at minute 52 of a 60-minute session, the median attendee has already left, not because they objected to the offer, but because they were gone before it existed.
Now the part that decides whether your follow-up can work at all. The exits are not random. They cluster. People leave around the agenda slide, around the first product mention, around the case study, around the transition into the pitch. Your platform records the minute. That record is a segmentation input, because it tells you which of four or five messages a person needs.
And here is the trap: you will only have that record after the event. If you wait for it, you have already missed the window that matters.
The Follow-Up Sequence, Written Before the Invitation
The advice you will find elsewhere treats follow-up as the work you do afterwards: send the replay within the hour, split attendees from no-shows, expire the replay at 72 hours, escalate. That advice is good, and the WebinarNinja guide ranking highly for this topic sets out a version of it I would happily run.
It also hides an assumption. Every one of those instructions is about acting on data you only have at the end. "Send the no-shows the replay" presupposes you can write that email on the day. In practice, the email has to clear review, the reviewer is in another time zone, the replay is still encoding, and the send lands on Thursday. That is not a follow-up. It is a newsletter about something that happened.
So the sequence has to exist before the invitation goes out. Four things get decided at that point, and none of them can be decided later without a delay you will not recover.
The segments. Pick them from the exit pattern you expect, not the one you observe. Four is usually right: attended to the offer, attended and left before it, registered and never came, clicked the offer. You will correct the boundaries afterwards. You cannot start without them, because each one needs its own copy.
The message each segment gets, in one sentence. Attendee who stayed: the thing you saw, and the next step. Attendee who left early: the minute you missed, with a timestamp. No-show: the honest acknowledgement that they did not come, plus the shortest version of the argument. Offer-clicker: not a follow-up at all, but a direct message from a human, because they told you what they wanted.
The order and the deadline. Replay up within the hour, the missed-minute email the next morning, the argument email on day two, and a firm close. A deadline is what turns a sequence into a decision. Without one, the last email sits in a queue and competes with everything else you send.
Who writes it. This is the decision that fails most often. If the follow-up is assigned to whoever is free after the event, you have not designed a sequence. You have scheduled an obligation. The person who wrote the invitation is the person who should write the follow-up, in the same sitting, because the promise in one has to match the payoff in the other.
That last point is the whole argument compressed. Your promotion made a promise. Your follow-up has to keep it. If the promise was a template, the follow-up can only offer another template. If the promise was the hour, the follow-up can offer the hour's argument, and the replay, and a conversation. Register more people with the first promise and you have grown a number while shrinking the room that could have bought something.
Measuring a Webinar Beyond Registrations
Registrations are the only webinar number that goes up when the work goes badly. So they cannot be the headline. Five numbers can carry the weight instead, and you need all five because each one catches a different failure.
What you measure | What it tells you | What a bad value means |
|---|---|---|
Attendance rate, by channel | Whether promotion selected for people with a reason to come | Your registration offer is buying the wrong crowd |
Median watch duration | Whether the room was still there when the argument arrived | You are asking for something after most people left |
Follow-up open and click rate, by segment | Whether the segmentation was real or decorative | You sent four emails to one audience |
Response rate from the offer-clicker segment | Whether anyone was ready to talk | Your offer did not land, or your topic attracted browsers |
Attendees per hour of team time | Whether the format is paying for the effort it consumes | The event is a cost centre with a nice slide |
Two of these, attendance by channel and median watch duration, need nothing new. Your platform already has them and nobody opens that tab. The fourth is the one to watch for a quarter before trusting it, because that segment is usually too small to read early.
One warning on the number everyone wants: pipeline attributed to a webinar is easy to inflate and impossible to defend in a room where someone is holding an attribution argument. Our breakdown of demand generation strategy settles how that credit should be allocated, and this page will not re-argue it. Keep the webinar's claim to what you can trace, and treat the rest as influence rather than revenue.
If you want a smaller, more honest set of numbers to report upwards, the discipline in our marketing KPIs piece applies here too: a metric whose bad value does not change your next decision is decoration.
When a Webinar Is the Wrong Format
A webinar is the wrong instrument more often than the people who run them admit. Three cases, and all three are common.
When you need breadth. If the goal is to reach a thousand people with a message, a recorded video and a paid distribution budget will do it for less. A webinar's advantage is depth of attention from a small group, and it evaporates at scale, which the 1,000-registrant attendance figure shows plainly.
When there is nothing to teach. A webinar with no teachable content becomes a sales call with an audience, and audiences leave those early. If the honest version of your topic is a product demo, run a demo. It will convert better and cost less.
When the follow-up cannot be resourced. This is the hardest one to accept. A webinar without a real sequence produces registrations and goodwill and nothing traceable, and the sequence is a repeated cost every quarter. If you cannot staff the four messages, do not run the event. Run a gated piece of content instead, and use a customer journey map to see where in the buying process the audience actually sits.
The third case is the one I would argue hardest about, because the registration number is exactly what gets used to defend it. The test I use now is simple: write the follow-up first, read it back, and ask whether anyone on that list would answer it. If the honest answer is no, the room is not the problem.
Frequently Asked Questions
What counts as a good webinar attendance rate?
Somewhere between 40 and 50 percent of registrants, depending on how warm the list is. Contrast's 2026 benchmark from over a million registrants gives 44 to 50 percent, and a separate 2026 summary puts the median at 41.6. Anything much above half often means you promoted to your own list, which is a good sign about the list and not a scalable channel.
Should you send the replay to people who did not attend?
Yes, and separately. A no-show registered on purpose, so they are warm, but they did not get the argument. The message that works acknowledges they missed it and offers the shortest version of the point rather than a bare link. If your replay email to no-shows is identical to the one you send attendees, you have skipped a segment.
Which webinar program should you pick?
The platform is a delivery decision, and it matters far less than the sequence you build on top of it. Three things are worth checking, and none of them is the interface: whether it records the minute each attendee left, whether you can trigger different emails per attendance segment without a developer, and whether the replay can be given a hard expiry date. Pick on those and the follow-up writes itself. Pick on templates and you will spend the quarter rebuilding the automation by hand.
How long should a webinar be?
Long enough to make the argument properly and short enough that people are still there when you make the ask. The 2026 data is blunt about this: average live viewing duration is 37 minutes, and 90-minute sessions hold attendance better than 30-minute ones but ask more of the calendar. Anywhere between 40 and 60 minutes is a fair default. Put the offer at the point where the median attendee is still in the room, not at the end because that is where offers usually go.
Write the follow-up before the invitation
The sequence is a design decision, and it is easier to make when the promotion, the landing page and the follow-up sit in one place. That is what Allable is for.