B2B Webinars

A Scheduled Talk, a Registration Form and a Recording

A webinar is a scheduled, single-session presentation delivered online, gated behind a registration form, and preserved as a recording that continues to be watched afterwards. Those three components are the product. The presentation is why people register, the form is the commercial mechanism, and the recording is usually where most of the value ends up.

A virtual event is a different product: multiple sessions, session-level rather than event-level attendance, different registration behaviour and a materially different cost structure. Benchmarks for one do not transfer to the other, and vendors blur the two constantly. A product demonstration is different again, and advertising one as the other loses the audience you have just paid to assemble.

What makes b2b webinars persist is not attendance. It is that a registration form is one of the few remaining contexts in which a business buyer will voluntarily supply their role, their company and an interest in a specific problem.

Every Published Attendance Figure Is Vendor Data

State this before quoting anything. Essentially every widely cited webinar performance statistic comes from a webinar platform analysing its own customers' events — a self-selected sample twice over, since only companies that bought that platform are included, and only the events they chose to run on it. There is no independent, cross-platform study of B2B webinar performance. No percentage on this page is offered as an industry fact.

The most-cited source is ON24's annual benchmarks report, drawn from its own platform. Its 2025 edition, covering 2024 data, claimed 57% registration-to-attendance and a split of 56% live and 45% on-demand. Its 2026 edition, covering 2025 data, claimed 60% registration-to-attendance, average attendance duration of 49 minutes and 239 attendees on average, with on-demand described as 50% of total attendees and live as 67%. Those last two sum to more than 100%, which tells you the definitions overlap. ON24 does not disclose how many webinars were analysed or the industry mix.

Treat movement between editions as a change in one vendor's client mix, not an industry trend, and ignore aggregator listicles, which cite each other.

Registration and Attendance Are Two Different Funnels

People register and attend for different reasons. Registration is a response to a promise; attendance is a response to a calendar.

What earns a registration:

• A specific promise rather than a topic. A session titled after a question the reader has actually asked outperforms one titled after a subject area.
• A named speaker the audience recognises — an internal expert with a track record, a customer willing to describe what happened, or an independent practitioner.
• A narrow question, because the reader can tell in one line whether it applies to them.
• Plain logistics: date, time in the reader's zone, and running time.

What earns attendance is different: reminder sequences, a calendar file that works, a start time suited to the audience's working day rather than the presenter's, and a promise credible enough to beat whatever else lands in that hour. The gap is structural — a registration is a commitment made days ahead by someone whose diary is not their own. Plan for a large no-show share and treat the registrant list as the asset.

The Recording Is the Product

On-demand viewing is a large share of total viewing and it accumulates after the live date. Every vendor report that splits live from on-demand finds the on-demand fraction substantial, and the consequences are operational.

Measure over a 60 to 90 day window rather than on the day: a same-day report captures a fraction of the audience and none of the pipeline.

Build the follow-up assets before the live session: the recording, an edited highlight of the best five minutes, a written summary for people who will not watch anything, the questions and answers written up as a page in their own right, and whatever the presenter promised on the day. Producing these a fortnight later misses the window in which the registrant list still remembers registering.

Decide gating deliberately. Gate the recording when the registrant list is what you want and the topic is narrow enough that the form is worth completing; leave it open when the material builds your reputation with people who will never fill in a form.

Co-Hosted Sessions and Partner Audiences

Co-hosting exists to borrow an audience. A partner with a list of the right kind of companies can double your registrations, and the arrangement fails for predictable reasons.

Agree the promotion commitment in writing before anything is scheduled: how many sends, on what dates, to which segment, plus social posts. The common failure is a partner who agrees enthusiastically, sends one email the day before, and delivers a handful of registrations while you carried the production.

Settle the data question before the form goes live. Each party needs its own lawful basis for the contacts it receives, each must tell recipients where the data came from, and the form has to disclose that both organisations will receive the details.

Choose partners for audience overlap without product overlap: complementary suppliers to the same buyer, a systems integrator, a trade association, or a customer willing to co-present. Content syndication networks that sell a registrant audience are a different transaction — the audience is paid rather than yours, and behaves differently on follow-up.

The Four Ways the Format Fails

Most disappointing webinars fail in one of four familiar ways, and all four are avoidable at planning stage.

• A product demonstration advertised as education. This happens because the person holding the budget wants pipeline this quarter and sees an audience sitting there. Attendees leave in the first ten minutes and do not register next time. If you want to run a demo, advertise a demo.
• Forty-five minutes of slides. Usually a deck built to be complete rather than watched. Replace the middle with something being shown: a live walkthrough, a real document, a worked calculation.
• A panel with no disagreement. Four people agreeing pleasantly for an hour teaches nobody anything, and it happens because panels get assembled from whoever was available and briefed to be positive. Recruit people who hold different views, and let the moderator ask which of them is wrong.
• No question answered. A session that surveys a subject without resolving anything leaves the audience where they started. Choose one question and answer it.

A related error is too many speakers: each adds handover time and rehearsal cost, and nothing the audience asked for.

What to Do With the Registration List

The registration list is the reason to run the session, so handle it as a process rather than an export.

1. Segment by behaviour: attended and stayed, attended and left early, registered and did not attend, watched on demand later. Those four groups warrant different messages.
2. Follow up within a day or two, while registering is a recent memory, with the recording and the promised material rather than a sales approach.
3. Route the ones who asked a question, requested contact or match your target account list to sales, with the question attached. A question asked in public is a stronger signal than any score.
4. Add everyone else to ordinary nurture, having told them at registration that you would.
5. Read the questions properly. They are the clearest statement of what your buyers do not understand.

What not to do is pass the whole registrant list to sales as leads. That is how a useful channel acquires a reputation for bad leads, and how a sales team stops working webinar registrants at all.

Measuring Over a Quarter, and When Not to Bother

Report a webinar over a 60 to 90 day window, and report audience composition rather than size. Useful measures: registrants matching your ideal customer profile or target account list rather than total registrants; attendance and duration as engagement signals rather than success criteria; on-demand views accumulating across the window; opportunities created from the cohort inside the window; and cost per qualified registrant.

The traps are specific. Comparing your attendance rate with a vendor benchmark measures you against a different client base using undisclosed definitions. Reporting on the day understates the channel systematically. Small numbers mean one large opportunity can make a mediocre webinar look excellent, so judge a series rather than a session. And self-reported source data rarely credits a webinar attended months earlier.

Do not bother when any of these is true: you have no audience to invite and no partner who does; the topic is really a demonstration; nobody has capacity to produce the follow-up assets, which is where the value is; or sales has no intention of working the registrants.

Frequently Asked Questions