Promote a webinar over six weeks, and plan for roughly half your registrations to arrive in the final seven days. Livestorm's 2026 Webinar Benchmark Report puts 49.6% of sign-ups in the last week and 15% on the day itself. The first four weeks build the audience that makes that last-week spike big enough to be worth having.
Below is the timeline ORRJO runs, the sends that go with it, the day-of-week data almost nobody acts on, and the numbers I could and could not verify while writing this.
How far in advance should you promote a webinar?
Six weeks. That is the window ORRJO publishes as phase three of its demand generation webinar process, and it matches what the registration data shows about when people actually sign up.
Two publishers have measured the curve and they do not agree with each other. GoTo, the company behind GoToWebinar, wrote in November 2017 that 69% of registrations arrive in the week leading up to the event, from an analysis of more than 300,000 webinars. In April 2019 the same company published 59% for the same measure. Livestorm's 2026 report, drawn from 33,000 sessions and more than seven million registrations, puts it at 49.6%, with 15% landing on the day.
Two of those readings are seven and nine years old and they contradict each other by ten points. I would not build a plan on any single one of them. What survives all three is the shape. The back half of the campaign does most of the work, and the day of the event is a real acquisition channel rather than an afterthought.
That shape gets misread. If half the registrations land in the final week, the tempting move is to compress the whole campaign into that week and save the effort. It does not work, because the last-week spike is a response to five weeks of exposure. Cut the run-up and you keep the shape and lose the volume.
The six-week timeline, week by week
Five columns, because a promotion plan with no owner column is a wish list.
| Week | What ships | Channel | Owner | Cumulative sign-ups |
|---|---|---|---|---|
| Six out | Topic locked, speaker confirmed, landing page live, LinkedIn event created, sign-up form asks the question you will answer live | Owned page, LinkedIn | Demand gen lead | 8% |
| Five out | Invitation one to the house list. Speaker posts it in their own words | Email, organic LinkedIn | Marketing ops, speaker | 18% |
| Four out | Personalised invitations to named target accounts, one to one, saying why this session and why this person | LinkedIn outreach, one-to-one email | SDR or campaign manager | 28% |
| Three out | The supporting asset. A short piece that stands on its own and carries the sign-up link inside it | Blog, newsletter, partner sends | Content | 38% |
| Two out | Invitation two to non-openers, different subject line, different angle on the same problem. Partner and speaker promotion. Paid, if it is running | Email, partner lists, paid social | Marketing ops | 50% |
| One out | Three sends: the agenda on Monday, a speaker post midweek, the join link the day before. Registrant list checked against the target account list | Email, LinkedIn | Marketing ops | 85% by the eve |
| Live day | Join link in the morning with the time in the reader's own zone. Host in the room fifteen minutes early. Questions pre-loaded from the sign-up form | Host and producer | 100% | |
| Week after | Recording to everyone who registered, with a reason to reply. Calls to the full sign-up list, attendees and no-shows alike. Meetings into diaries | Email, phone | Sales | Registrations closed, meetings start |
Read the last column carefully, because it is two things stacked. The final-week and day-of figures come from Livestorm's 2026 report. The split across weeks six to two is ORRJO's own planning shape, because no publisher I could find breaks the early weeks out at all. Treat those five numbers as the ranges we plan against, not promises, and replace them with your own curve as soon as you have run three sessions.
The owner column is the part most plans leave blank, and it is where webinar campaigns quietly die. A landing page with nobody's name against it in week six becomes a landing page that goes live in week four, which deletes a third of the campaign before anyone notices. In-person events run on a different clock and a longer one, and that timeline sits in B2B event lead generation.
How many emails should a webinar invite sequence have?
Six to the house list, arranged so four of them land in the last eight days.
- Week five, the invitation. Topic, speaker, date, what the attendee walks away with.
- Week two, the resend. Same session to people who did not open, different subject line, a different angle on the problem.
- Week one, Monday, the agenda. What the twenty-five minutes actually covers, in order.
- Week one, the day before. The join link and the running order.
- Live day, morning. The join link again, with the start time in the reader's own time zone.
- Week after, the recording. To everyone who registered, attendees and no-shows.
Email is doing almost all of the work here. Livestorm's 2026 report puts 86.3% of registrations through email and has 20.4% of teams using paid social to drive sign-ups. Goldcast's growth team, writing in February 2023 about its own programme, described sending "between two to four emails before a webinar, depending on the scale of the event".
What nobody publishes is a controlled test of how many reminders lift attendance. Figures circulate. Three reminders lift attendance 28%. A one-day, one-hour and five-minute cadence lifts it 27%. I opened the pages carrying both and each one cites another blog, which cites another blog. No primary source publishes this. Until somebody runs the test and shows their working, treat the reminder count as a judgement call, and price in that every extra send is an invitation to unsubscribe.
What a good webinar invitation says
The structure has five parts and the order matters more than the wording.
- The problem, in the reader's language, in one line.
- What the session covers, in specifics rather than themes.
- Who is speaking and why they would know.
- The cost in time, with the Q&A counted inside it.
- The link.
Here is that structure filled in. It is illustrative, and the two bracketed slots are the parts you supply.
Subject: The tender question your carriers keep asking
Most logistics teams find out what a contract was really worth about six months after they sign it. [Speaker, title, company] has run that process from both sides and is doing twenty-five minutes on it on Thursday 6 August, with five minutes of questions at the end. Covered: how to price a lane before you can see the volume, and the two clauses that decide whether a rate holds when fuel moves. If it is useful to you, you will know inside the first ten minutes. [Register here]
Two things that example does without. There is no company boilerplate, and no adjective is doing a job that a fact should be doing. The reader is deciding whether to hand over half an hour of a working day, and the only argument that lands is a specific description of what happens inside it.
One more thing about the subject line. Goldcast read 900 webinars across 100 B2B brands and reported in February 2023 that using the word "webinar" in the title was "associated with a 50% drop in registration". Association is not cause, and 900 events is a small sample next to the platform-wide reports. It is still the cleanest reason I have seen to describe the session instead of the format.
How do you get more people to attend?
Move the date before you touch the copy.
Goldcast's 2026 B2B Webinar Benchmark Report covers 26,190 webinars run by 522 B2B organisations in 2025, 6.8 million registrations and 2,220,718 attendees. Goldcast presented it publicly on 23 June 2026. The day-of-week table inside it is the most useful thing published about webinars this year, and hardly anybody acts on it.
| Day | Share of B2B webinars hosted | Attendance rate |
|---|---|---|
| Monday | Not published separately | 47.2% |
| Tuesday | 24.42% | 41.2% |
| Wednesday | 30.53% | 39.0% |
| Thursday | 30.03% | 38.7% |
| Friday | Not published separately | 46.9% |
Source: Goldcast, 2026 B2B Webinar Benchmark Report. Tuesday, Wednesday and Thursday account for 84.98% of the sample between them, which leaves roughly 15% split across Monday and Friday. The report does not break those two days out by volume, so I have left the cells empty rather than fill them.
The three days that host 85% of B2B webinars are the bottom three on attendance rate. Monday converts eight points better than Wednesday. Goldcast's own summary line is that "Monday has the highest attendance rate of any weekday, and Friday afternoon outperforms".
The fair objection is that Monday draws a smaller list, so a better rate on fewer registrations may not put more people in the room. The report does not publish registrants per day in a form I could verify, so I am not going to pretend the trade is settled. For a monthly programme, alternate Monday and Wednesday for two quarters and you will learn more from your own numbers than from anyone's benchmark.
On the overall attendance rate, the two large 2026 datasets disagree. Goldcast puts it at 40% across its 26,190 webinars, up from 33% in 2024. Livestorm puts it at 51.3%. Both are platform data, both are honestly reported, and they measure different customer bases doing different things. Anyone quoting one of them as the webinar attendance rate is quoting their own users. The wider spread of published figures sits in B2B webinar benchmarks.
Two other levers, both cheap. The first is length. Goldcast found a median duration of 60 minutes, with 50.5% of sessions running exactly an hour. ORRJO runs thirty, with twenty-five minutes of content and five to ten for questions, published on the webinars and events page. An hour is a meeting somebody has to move something for. Half an hour is a coffee.
The second is the recording. Goldcast found 89.1% of webinars are made available on demand, that 91% of on-demand viewers watch nearly the whole thing, and that live attendees complete 74%. The people who watch it later finish more of it than the people who turned up.
What is a realistic registration target?
Goldcast's 2026 average is 251 registrants per event across 522 B2B organisations, up from 238 in 2024, with an average of 102 attendees.
That average hides an enormous spread and it is close to useless as a target for a first session. No report I have read breaks registrations down by the size of the list that was invited, which is the variable that decides the number. Work from your own list instead. Take the share of a warm house list that has ever signed up for anything you have run, apply it to the names you can reach this time, and you have a target you can defend in a planning meeting. 251 is not that.
For scale, two ORRJO programmes with published numbers. A webinar series for Jua, a Swiss AI weather prediction company, produced 556 webinar sign-ups. The Aveni demand generation programme, now in its fifth year, has produced 1,000 webinar sign-ups and 500+ meetings.
Registration volume against attendance quality
This is where webinar programmes go wrong, and it is a targeting decision rather than a promotion one.
A registration number is easy to move. Widen the audience and soften the topic, and 400 people will sign up for a session about the future of an industry. Then 40% attend. A good share of them are competitors and consultants, and the follow-up list is worth almost nothing to the sales team who inherit it.
Take two sessions with identical follow-up. The broad one draws 400 registrations and 160 attendees at Goldcast's 40% rate, of whom perhaps a quarter sit in the function you sell into. That is 40 names worth a call. The narrow one draws 90 registrations and 46 attendees at Livestorm's 51.3%, and most of them are the people you sell to. Same callable list off a fifth of the registrations. The quarter is my illustration rather than a measured figure, and it is exactly the number your planning meeting should be arguing about.
The reason ORRJO measures a webinar programme at the meeting rather than the registration is that a meeting is where the number stops being flattering. Across 12,500+ meetings booked, ORRJO reports a 90%+ meeting attendance rate. A meeting that gets attended came from somebody who had a reason to be there, and that reason was set six weeks earlier, when you chose the topic and decided who to invite.
Four things we stopped doing
- Leading with the word webinar. Goldcast's 2023 reading is correlational and I would not lean on it alone. The underlying point holds without it: the format is no reason to give anyone half an hour of your day. Describe the session.
- Defaulting to Wednesday. The Goldcast day-of-week data is the reason. Wednesday is the crowded slot and the worst-converting one of the five. Booking it was habit rather than a decision.
- Twelve-week runways. A campaign that long decays in the middle. The material goes stale and the speaker loses interest. The sends in weeks twelve to seven earn close to nothing against the effort. Six weeks is the published window and it holds up.
- Sending the recording as a courtesy. Every recording now goes out with a specific reason to reply, and the whole sign-up list gets worked by phone in the week after. ORRJO sells that call-down as its own line item, a post-webinar sales day, which tells you how much of the value sits after the event rather than during it.
Before you write a word of copy
Do two things. Pick the date off the attendance data rather than the calendar convention everybody else follows, and write a name against every row of the timeline above. Most of what goes wrong in a webinar campaign is recoverable in week two. Those two are not.
ORRJO runs webinar programmes inside demand generation, from topic research through to the follow-up calls, from £2,500 a month. If you would rather compare the market first, we wrote the honest list: best B2B webinar agencies in the UK. Or book a strategy call and we will look at your next session.