What’s a good webinar conversion rate? As of October 2026, only one stage of the funnel has large published benchmarks, registrant to attendee, and three large platform reports disagree: Goldcast puts the average at 33%, Livestorm at 47.7% and ON24 at 60%. Past attendance, those reports publish engagement data but no webinar-specific rates for attendee to MQL, MQL to opportunity or opportunity to closed-won, so for those stages you need a baseline built from your own CRM.
I’ve produced more than a thousand enterprise B2B webinars, and my rule for funnel numbers is simple: use a published benchmark where one exists and say whose it is, and use your own trailing baseline everywhere else. This playbook gives you both, plus what to check first when a stage comes in low.
The full funnel at a glance
Quick verdict first: what has been published for each stage as of October 2026, and what to fix first when your own number drops below your baseline.
| Funnel stage | Published benchmark (as of October 2026) | If you’re below your baseline, check this first |
|---|---|---|
| Registrant to attendee | Goldcast: 33% average (2025 report). Livestorm: 47.7% show-up (2025 data). ON24: 60%, counting live and on-demand viewers (2025 data). | Reminder sequence and time slot |
| Live attendee to engaged attendee | No shared definition. Livestorm: 15.4% active participation in 60 to 75 minute sessions; average watch time 26 minutes. Goldcast: average watch time about 29 minutes. | Format and interaction cadence |
| Engaged attendee to MQL | No webinar-specific benchmark found. Measure your own. | Scoring model and follow-up speed |
| MQL to SQL or opportunity | No webinar-specific benchmark found. Compare against your other lead sources. | Sales handoff context |
| Attendee to pipeline and closed-won | No webinar-specific benchmark found. Track each event’s attendees through a full sales cycle. | Everything above, compounding |
Two definitions first. In this playbook an “engaged attendee” stayed 30 minutes or more, or interacted at least once: a poll response, a question, a resource download. That’s a high bar, since the average viewer stayed 26 minutes on Livestorm in 2025 and about 29 minutes on Goldcast, so set your threshold against your own session length. “Qualified pipeline” means opportunities sales accepted and is actively working, not everyone who showed up.
Registrant to attendee: the published numbers
This is the one stage with large public datasets, and they don’t agree, because each covers a different platform’s customers and counts attendance its own way:
- ON24: 60%. ON24’s 2026 Digital Engagement Benchmarks put the average registration-to-attendee rate at 60% for 2025. ON24’s glossary defines that rate as registrants who attend a live or on-demand event, so replay viewers count.
- Livestorm: 47.7%. Livestorm’s 2026 benchmark report covers 33,786 sessions from 3,199 organizations in 2025 and puts the show-up rate at 47.7%, down from 48.9% in 2024. It leaves out sessions with fewer than 10 registrants or shorter than 15 minutes.
- Goldcast: 33%. Goldcast’s 2025 B2B Webinar Benchmark Report analyzed 19,531 webinars across 418 B2B companies (2024 data) and gives an average attendance rate of 33%, up from 29%. Goldcast’s newer 2026 report covers 26,190 webinars from 522 companies, but its public summary gives only totals (6.8 million registrations, 2.2 million attendees) and no average rate.
Don’t average these into one number. Compare yourself with the report whose counting matches yours: Livestorm and Goldcast if you count only the live room, ON24 if your platform counts replay viewers as attendees. And segment by audience first. In my view, a registrant from a cold paid click shouldn’t be judged against a subscriber who has read your newsletter for three years.
If you’re well below the report that matches your setup, check these first:
- Your reminder sequence. Four emails is my working standard: the confirmation, a value email three days out, the day-before reminder, and an hour-before email with the join link at the top. Livestorm reports that teams with show-up rates above 60% send personalized reminders three days out, the day before and the day of. The full cadence is in our four-email reminder sequence.
- Your slot. Tuesday through Thursday, 11am to 2pm ET, is the window I’d pick, but that’s my preference, and the data doesn’t agree on a best day. Livestorm’s 2025 data has Tuesday on top at 51.7%, with every weekday between 50.3% and 51.7%, and it recommends Monday to Wednesday at 11am or 2pm ET, avoiding the 12 to 1pm lunch hour. Goldcast’s 2025 data has Monday highest at 38.21% (Tuesday 34.10%), and tech companies peaking on Fridays from 12 to 3pm at 44%; Goldcast doesn’t state a time zone. Whatever you pick has to survive your audience’s calendar: in Zoom’s 2025 audience research, 68% blamed no-shows on a schedule conflict. Test a new slot against your trailing average before you commit to it.
- The gap between registration and the event. Registrations arrive late: Livestorm found 49.6% came in the final week and only 16.4% three or more weeks out, and Zoom reports a 13-day median promotion window for the highest-attended webinars on its platform. I’d still open registration three to four weeks out for a flagship, but give the early registrants a value touch mid-sequence (a teaser clip, a one-question poll), and keep registration open until the start.
Attendance by audience source, day and format gets its own page: average webinar attendance rates.
Live attendee to engaged attendee
This is the stage I’d watch most closely, because it feeds every stage after it. The platforms publish engagement data, but none of it is an engaged-attendee rate you can copy:
- Livestorm (2025): average watch time of 26 minutes against a 68-minute average session. Active participation was 15.4% in 60 to 75 minute sessions, rising from 12.5% for 30 to 44 minutes to 20.7% for 90 minutes or more.
- Goldcast (2025 report): average watch time of about 29 minutes; per webinar, an average of 24 chats, 17 poll submissions and 6 questions.
- ON24 (2025): 1.8 interactions per attendee.
Two things in that data are worth acting on. Polls are underused: Goldcast’s 2025 report lists poll usage at 15%, its 2026 report says polls appear in only 19% of webinars, and Livestorm puts poll usage at 28%. And Livestorm’s numbers don’t show shorter sessions engaging better, since participation rose with length. Livestorm doesn’t say whether longer sessions cause that or attract more committed audiences.
My own rules for this stage: keep a top-of-funnel session to a tight 45 minutes (my preference; no dataset here backs it), put an interaction point (a poll, a chat prompt, a show-of-hands question) every 8 to 10 minutes, and take questions throughout so the people with the strongest intent have a reason to stay. If your engaged share falls below your trailing baseline, look at the content design before you blame the audience.
Engaged attendee to MQL: measure your own
I couldn’t find a published, webinar-specific benchmark for this stage from a platform vendor or research firm as of October 2026. The figures that circulate on marketing blogs don’t trace to a published platform or research dataset I could open, and MQL definitions vary so much between companies that a borrowed number would mislead you anyway. Build the baseline yourself:
- Write the definition down before the event: your MQL threshold and your window. I’d use 14 days from the event for live attendees and 14 days from the view for replay viewers.
- Divide new MQLs by engaged attendees, and also by all attendees, so events with different engagement stay comparable.
- Remove people already in play. Anyone who was an MQL, an open opportunity or a customer when they registered comes out of both counts.
- Wait for the window to close, then baseline against your trailing four to six events.
A worked example, with made-up round numbers: 200 engaged attendees, 20 of whom were already MQLs, open opportunities or customers when they registered, leaves 180. If 45 of those 180 cross your MQL threshold within 14 days, that event’s engaged-attendee-to-MQL rate is 45 ÷ 180 = 25%.
If an event comes in below your baseline, check two things in order. First, your scoring model. If it awards a flat “attended webinar” score and ignores behavior (minutes watched, questions asked, polls answered), your hottest attendee scores the same as someone who joined for four minutes on mute. Our webinar lead scoring model walks through a full point structure. Second, follow-up speed. My rule is a first follow-up within 24 hours, segmented by behavior; a generic “thanks for attending” blast on day four wastes that window. The segmented alternative is in our follow-up playbook.
MQL to SQL and opportunity
No published, webinar-specific benchmark turned up for this stage either. The comparison that matters is internal: webinar-sourced MQLs against MQLs from your other sources (content downloads, paid search, events) in the same CRM, with the same definitions and window. If you don’t have a window yet, I’d start at 90 days from the MQL date for long B2B cycles and keep it fixed.
When webinar MQLs convert worse than your other sources, I’d look at the handoff before blaming lead quality. Sales gets a name and a lead source, calls, and opens with “I saw you attended our webinar.” The fix is routing context with the record: the question they asked in Q&A, verbatim; their poll answers; how long they stayed. A rep who opens with “you asked about migration timelines on Thursday, want to go deeper on that?” is having a second conversation, not a first one. The routing mechanics are in our webinar-to-sales handoff playbook.
Attendee to pipeline and closed-won
Compound the stages and you get the number executives ask about. I found no published webinar-specific benchmark for attendee to pipeline or attendee to closed-won, so measure your own:
- Pipeline rate = sales-accepted opportunities tied to the event ÷ total attendees, read at a fixed window. I’d use ninety days, applied the same way to every event, so quarters stay comparable.
- Closed-won rate = closed-won deals from the same attendees ÷ total attendees. Don’t read it at 90 days; wait at least one median sales cycle from the event, or long deals will be missing from the count.
The common blocker is plumbing. Livestorm reports that only 6 to 8% of its customers use native CRM integrations (Salesforce, HubSpot, Brevo) to sync webinar data to their pipeline. If attendance and engagement never reach the CRM, none of these rates can be computed.
Two habits I’d add wherever your numbers land. Work the replay audience as hard as the live one (next section). And run series instead of one-offs: in Goldcast’s 2025 data, webinars with “series” in the title averaged 309 registrations and a 38% attendance rate, against its benchmarks of 238 and 33%. Turning pipeline into a defensible ROI figure is its own discipline, covered in webinar ROI and pipeline attribution.
Replay and on-demand viewers
How big the replay audience looks depends on who’s counting. ON24 reports that 67% of attendees joined live and 43% watched on demand in 2025, so some did both. Goldcast’s 2025 report measures against registrants and lands far lower: 9.5% of registrants watched on demand, 7.15% watched only on demand, and 31% of live attendees came back to rewatch. Its 2026 summary adds that 91% of on-demand viewers watch nearly the entire recording. Access matters too: Livestorm found public replays averaged 14.60 on-demand views, against 3.97 for replays open only to registered users.
What I couldn’t find is published data on whether replay viewers convert to MQL or pipeline at the same rate as live attendees. So measure it: score replay engagement the same way as live (minutes watched, resources clicked) with a modest discount, follow up on the same 24-hour clock from the view, and report replay conversion separately before you combine it with live.
Measuring your funnel honestly
For every stage after attendance, your own trailing numbers are the benchmark. This is the worksheet I’d set up; the windows are my recommendations, and none of them is a published standard.
| Stage | Formula | Window | When to read it |
|---|---|---|---|
| Registrant to attendee | Live attendees ÷ registrants; on-demand viewers reported separately | Event day | After each event |
| Live attendee to engaged attendee | Engaged attendees ÷ live attendees | Event day | After each event |
| Engaged attendee to MQL | New MQLs ÷ engaged attendees, minus people already in play | 14 days | When the window closes |
| MQL to SQL or opportunity | SQLs or opportunities ÷ webinar-sourced MQLs | 90 days, or your standard window | When the window closes |
| Attendee to pipeline and closed-won | Sales-accepted opportunities, then closed-won deals, ÷ total attendees | 90 days for pipeline; one median sales cycle for closed-won | Quarterly |
Three rules I’d enforce: never move the attribution window to flatter a quarter. Cohort by audience temperature, because blending a customer webinar with a cold-list webinar produces an average that describes neither. And never let one blowout event set the baseline.
One more check. If one stage is far above your baseline while the next is far below, suspect the definition before you credit the team. When nearly everyone counts as engaged but few reach MQL, your engagement definition is too loose.
Frequently asked questions
What is a good webinar conversion rate?
It depends on the stage, and only the first stage has published benchmarks. For registrant to attendee, Goldcast reports a 33% average, Livestorm 47.7% and ON24 60%, with ON24 counting on-demand viewers. For the later stages no published webinar-specific benchmark turned up, so a good rate is one that beats your own trailing four to six events.
How do you calculate webinar conversion rate?
Pick the stage and divide the later count by the earlier count. Attendance rate = live attendees ÷ registrants. MQL rate = new MQLs within your window ÷ engaged attendees. Pipeline rate = sales-accepted opportunities ÷ total attendees. Fix your definitions and time window in advance and apply them identically to every event.
What percentage of webinar attendees become customers?
I couldn’t find a published, webinar-specific figure for attendee to closed-won, so treat any number quoted without a named dataset as a guess. Measure it on each event’s attendees after at least one median sales cycle has passed, once attendance data reaches your CRM.
Do replay viewers convert as well as live attendees?
No published comparison turned up, and even the replay audience’s size depends on who counts it: ON24 says 43% of attendees watched on demand in 2025, while Goldcast found 9.5% of registrants did. Track replay conversion separately until you have your own answer.
This is lesson-one material from our courses: see what’s coming, or get the free lessons by email as they ship.
Last checked: October 2026. The benchmark reports cited here are updated yearly; check the linked report before you quote a figure.