Webinar Fatigue Is Real: ON24 Sold for $400 Million and the Budget Moved On

Webinar fatigue stopped being an opinion in December 2025. Cvent agreed to buy ON24, the largest listed pure-play webinar company, for about $400 million in cash. That is $8.10 a share against a $50 initial public offering price in February 2021. The demand did not vanish. It split into two formats, and the 60-minute gated session sits between them.

What webinar fatigue actually describes

Webinar fatigue is usually described as an audience problem. People registered, people stopped showing up, and marketers concluded that everyone got tired of video calls.

That explanation is too flattering to the format. The webinar was never mainly a content product. It was a trade: your email address and 60 minutes of your calendar, in exchange for information you could not get any other way.

Both sides of that trade repriced. The information became easy to get elsewhere, and the calendar became more expensive to give away. When a trade stops clearing, the volume falls, and no amount of better production fixes it.

Registration is a price, and the price went up

A registration form is a payment. The buyer pays with contact details and a future slot in the working day, then waits days or weeks for delivery.

In 2021 that was a fair price, because the alternative was a sales call. In 2026 the alternative is an interactive demo, a public transcript, or a model that will summarise all three in nine seconds. The same price now buys much less relative advantage.

This is the same mechanism that reshaped the economics of the gated form. The gate did not get worse. Everything outside the gate got better.

The 60-minute slot is the part that broke

The duration was never chosen for the audience. It was chosen because it fits a calendar block and produces enough recorded minutes to justify the production cost.

That length made sense when the recording was the only artefact. It stopped making sense once the recording became raw material for clips, transcripts, summaries and answers. A format that costs the audience 60 minutes and delivers 6 minutes of new information is not tired. It is mispriced.

The clearest evidence is a cash offer, not a survey

Marketing debates about format decline usually run on survey data. This one has something better. A public company built entirely on the webinar sold itself, and the price is a matter of record.

From $203.6 million to $139.3 million

ON24 was the closest thing the category had to a pure play. It reported total revenue of $203.6 million in 2021, up 30% year over year, which was the high-water mark for the whole virtual-event boom.

Then it declined every year that followed. Revenue was $163.7 million in 2023 and $148.1 million in 2024, then $139.3 million in 2025. Total annual recurring revenue finished 2025 at $124.0 million, down from $129.7 million a year earlier.

Read the ARR line rather than the revenue line. Revenue is partly a lagging contract artefact. ARR is what customers have agreed to keep paying, and it fell in a year when the company reported its best gross retention in three years.

2022 is deliberately absent. Aggregator sites carry a figure for it, and this post lists only years checked directly against ON24's own results releases. The 32% fall is calculated from 2021 to 2025, both of which are verified.

Cvent bought two webinar platforms in 16 days

Here is the part that gets misread. The buyer was not a software conglomerate hunting for cash flow. It was Cvent, a company whose core business is running physical meetings and events.

Cvent announced its acquisition of Goldcast, an AI video and webinar platform, on 15 December 2025. It announced the ON24 agreement on 30 December 2025 at $8.10 a share, a 62% premium to the 10 November 2025 close. That deal closed on 1 April 2026.

Two webinar platforms, 16 days apart, absorbed by an in-person events company. That is not a vote of confidence in the standalone webinar. It is a judgement that the webinar survives only as a feature inside a wider event programme, next to the rooms and the badges and the first-party data.

Why webinar attendance is the worst-sourced number in B2B marketing

Search for a webinar attendance benchmark and you will find a tight cluster of figures. Roughly 40% to 50% of registrants attend live. Another 10% or so watch a replay. The numbers appear in dozens of posts and look settled.

They are not settled. Follow the citations and they collapse into a small number of platform vendors reporting statistics about their own customers, restated by aggregators that did not check. A statistic that appears in six places is still one statistic if all six trace to the same unpublished dataset.

Whose funnel is being measured

Platform benchmarks are drawn from the accounts that keep paying for the platform. That population is biased toward teams for whom webinars still work, because the teams they stopped working for churned out of the dataset.

So the published attendance rate is a survivor statistic. It can stay flat for years while the underlying practice shrinks around it, which is precisely what ON24's ARR line suggests happened. This is the same measurement gap covered in the piece on measuring discovery you cannot see in the CRM.

My position, stated plainly: I do not believe anyone has published a defensible longitudinal attendance series for B2B webinars, and I would not plan a budget on the ones in circulation. The audited numbers point one way and the vendor numbers point the other, which usually means the vendor numbers are measuring a shrinking sample well.

The format did not die. It split in two.

The useful frame is not decline. It is separation. Two things a webinar tried to do at once have moved apart, and each half now has a better home.

At one end, buyers still want the thing a room gives you: unscripted conversation, peer signal, and a reason to be somewhere. That is expensive per head and it is holding up. Gartner's 2026 CMO Spend Survey was fielded January to March 2026 among 401 marketing leaders. It found B2B respondents putting a mean 27.6% of offline marketing budget into events, well ahead of the 16.5% reported by B2C peers.

At the other end, buyers want the answer with no meeting attached. Gartner's sales research puts 67% of B2B buyers preferring a rep-free experience and 45% using AI during a recent purchase. Those buyers are not going to fill in a form and wait until Thursday.

The 60-minute gated virtual session is caught in between. It costs the audience like a live event and delivers like a document. That is the whole of webinar fatigue, and it explains why how buyers now assemble a shortlist matters more than how many people joined your call.

What B2B teams are running instead

Four formats are absorbing the budget. None of them is a webinar with a shorter runtime, which is the mistake most teams make first.

The four formats taking the budget

Each one picks a side of the split deliberately. That is the design decision, not the production value.

The interactive demo is the one with the clearest published evidence, and it comes with a caveat. Navattic's State of the Interactive Product Demo 2026 analysed more than 40,000 demos built on its own platform. It found that 66% of the top-performing demos were ungated, with higher engagement and completion than gated equivalents. That is the vendor measuring its own customers, so treat it as directional rather than settled.

The teardown works for a reason that has nothing to do with duration. It replaces a presentation with a demonstration of judgement. That is the same asset described in original research as a content moat, delivered live instead of as a PDF.

What none of these replace

All four are worse than a webinar at exactly one job: producing a large list of names on a fixed date for a quarterly number. If that is what your webinar programme is for, say so out loud before you replace it, because the replacements will look like failures against that metric.

The physical end of the split is genuinely holding. UFI's Global Exhibition Barometer covers 378 companies across 57 countries and was fielded to December 2025. It found 44% of organisers expecting activity to rise more than 5% in 2026, against only 8% expecting a decline of that size. Rooms are not the problem.

Where this argument is weakest

Three genuine problems with everything above, stated before someone else states them.

One company is not a category

ON24's decline is real and audited, and it is still a single company. It faced direct competition from Zoom and Webex, both bundled into software that buyers already pay for. Some of that revenue did not leave the webinar. It moved to a cheaper webinar inside a suite, where nobody reports it as a separate line.

If that is the dominant effect, then webinar volume may be flat or rising while the market value of standalone webinar software collapses. My reading of the ARR and retention data is that both things happened. I cannot prove the split between them, and anyone who claims they can is guessing.

The rep-free buyer still wants a rep

The 67% figure is regularly used to argue that human-led formats are finished. Gartner's own follow-up research undercuts that. In a survey published in May 2026, 69% of B2B buyers said they turn to sales representatives to validate AI-generated insights.

So the buyer wants to research alone and then check with a human. A live session is one of the cheapest ways to be that human at scale. That is the strongest available case for keeping a webinar, and it is a good one.

The replacement evidence is vendor evidence

Every figure in this post describing the decline comes from audited filings or a named analyst survey. Almost every figure describing the replacements comes from vendors selling those replacements. That asymmetry is not accidental, and it should lower your confidence in the second half of the argument relative to the first.

Replace the metric before you replace the format

Most format changes fail in month two, when someone pulls the demand generation dashboard and the registration count has dropped by 80%. The dashboard is measuring the old trade, so it reports the new one as a collapse.

Change what you report first, run the old format for one more quarter alongside it, and let the two be compared on the same terms. This costs one quarter and prevents the whole change being reversed on a bad Tuesday.

The last row is the one that changes behaviour. Once a dinner and a webinar are priced against the same denominator, most teams discover their cheapest format was not the one with the biggest number attached to it. Pair that with case studies with verified outcomes and comparison pages that buyers actually read, and the ungated half of the programme starts carrying its own weight.

One warning about volume. Replacing one webinar a month with twelve short assets is not a strategy, it is the volume economics of the content flood in a new costume. Fewer, harder, more specific pieces beat more of anything.

Frequently asked questions

How to test this in one quarter

Pick your next scheduled webinar and split it in half rather than cancelling it.

Run the live session as a 20-minute teardown of one real problem, with 15 minutes of unscripted questions and no slide deck. Publish the recording and the full transcript with no form in front of either. Then take the budget you would have spent promoting it and buy dinner for 15 people in your best city.

Report both against cost per meeting held, not cost per registration. Ninety days later you will have a defensible answer for your own market, which is worth more than every benchmark quoted in this post.

Related reading