How to Use Video for Lead Generation (by Funnel Stage)
How to use video for lead generation at every funnel stage, when to gate a video and when not to, the CTAs and forms that convert, and how to turn views into pipeline.
I have watched more video lead-gen budgets leak from one mistake than from any other, and it is always the same mistake: a single kind of video gets pointed at the whole buyer journey and asked to carry all of it. A fifteen-second social hook and a three-minute customer story do opposite jobs. When leads are not showing up, nine times out of ten I find someone asking one of those two to do the other one’s work. I co-founded Moonb, I have sat through a lot of lead-gen scoping calls, and the first thing I say on almost every one of them is close to boring in how simple it is: match the video type to the funnel stage. Everything below is how I actually split that up.
If you only want the playbook:
- Keep the top of the funnel ungated. Reach is the job there, not capture.
- Gate one or two high-value middle assets behind a short form, and gate per asset, never as a blanket policy.
- Put proof and personal video near the decision, where a real customer or a real face pulls someone off the fence.
- Wire watch percentage into your CRM so a view can turn into a routed lead instead of a vanity number.
Top of funnel: attract, do not try to capture yet
At the top I want reach and demand, so I keep these videos ungated and give them one low-friction next step. Short social clips, explainers, and thought-leadership pieces live here, and the second you put a form in front of one you have kneecapped it. A gated explainer is invisible to Google and to AI answers like ChatGPT and Perplexity, it earns no links, and it never compounds into the traffic that fills the rest of the funnel. I have talked more than one client out of gating their best awareness video, because the handful of “leads” a gate captures this early is a rounding error against the reach it costs you. Leave the front door open, let the video do demand generation, and ask for the email later when the person has an actual reason to give it.
Middle of funnel: qualify with demos, webinars, and comparisons
In the middle, intent is higher and gating starts to earn its place, but I gate per asset and I decide it one video at a time. Product demos, comparison videos, and webinars are worth a form, because the person watching is evaluating for real. What I care about far more than whether to gate is where the gate sits: a wall before any playback tanks completion, while a mid-roll turnstile that appears after I have already delivered a hook of real value converts far better. My rule to clients is short. Deliver something worth watching, then ask. On a webinar I would rather register people up front and gate the on-demand replay; on a two-minute demo I would let the first thirty seconds run clean and drop the form once the value is obvious.
Bottom of funnel: close with proof and a personal video
Near the decision, two formats do most of the lifting, and I lean on both.
- Testimonials and case studies. A real customer naming a specific, measurable outcome (“cut onboarding time 40 percent”) beats generic praise every single time I have tested it. I place these next to pricing and decision points on the page, not buried at the bottom where nobody scrolls.
- Personalized 1:1 sales video. A webcam or screen-share recorded for one named prospect is the highest-leverage bottom-funnel move I know. Authenticity beats polish here, so I keep it under sixty seconds and hold it to a single idea. A custom thumbnail earns the click (your name written on a whiteboard works fine), and the video should end on one clear ask and nothing else. Adding video to outreach measurably lifts replies, and Vidyard’s own benchmarks put personalized sales-video reply rates in the low double digits.
The by-stage cheat sheet
When a client asks me to put the whole thing on one page, this is roughly the table I hand over. It maps the stage to the video type, tells you whether I would gate it, and names the one CTA I would attach.
| Funnel stage | Best video types | Gate it? | The one CTA | Length I aim for |
|---|---|---|---|---|
| Top (attract) | Short social clips, explainers, thought-leadership | No | Grab a free guide or checklist | 15 to 60 seconds |
| Middle (qualify) | Product demos, webinars, comparison videos | One or two, per asset | Register or book a demo | 2 to 10 minutes |
| Bottom (close) | Customer testimonials, case studies | No | Book a 15-minute call | 60 to 90 seconds |
| Bottom (close) | Personalized 1:1 sales video | No | Reply, or grab a time | Under 60 seconds |
The pattern that runs through the whole table is that I only gate in the middle, and even there I gate the exception, not the rule. If you want the mechanics of shooting one video and versioning it for every one of these placements, I wrote that up separately in best video formats and aspect ratios.
Forms and CTAs that convert
- One CTA per video, specific to the stage. Awareness asks for a guide, consideration asks for a demo or a webinar seat, decision asks the viewer to book a fifteen-minute call. I write these as plain imperatives and I never end a video on “let me know your thoughts,” which asks for nothing and gets nothing.
- Short forms. Field count moves inversely with completion, so I ask only for what I will actually use to route or qualify, which usually means an email and sometimes a name and company. When a client wants more, I use progressive profiling to collect it across later touches rather than stacking a long form on the first ask.
Turn views into pipeline
A view is not a lead until your systems treat it like one, and this is the part teams skip and then wonder why the video “did not generate anything.” I connect the video platform to the CRM and to marketing automation so watch percentage triggers action. Someone who gets through 50 to 75 percent of a demo gets flagged to sales or dropped into a nurture sequence automatically. That engagement data is the qualification signal, and it is what separates real video lead-gen from a pile of view counts nobody can act on.
A few wrapper details compound the whole effect once the routing is in place. The landing page has to be fast and mobile-first, because most plays happen on a phone. Captions with silent autoplay matter, since feeds run on mute by default. And the first three seconds carry a real hook, because that is exactly where most of the drop-off happens. If you want to push further, interactive video with branching paths and in-video questions turns a passive viewer into a self-qualified lead and captures richer intent than a form ever will.
What I would actually do first
If you match the video to the stage, gate only the one or two assets that earn it, ask for a single thing, and let engagement data route the lead, video stops being a view counter and starts behaving like a pipeline engine. At Moonb I build video for each stage of the funnel as an embedded creative team, wired to convert rather than just to look good. If lead-gen is where your video should be pulling its weight and it currently is not, see how Moonb works.
Frequently asked questions
Both work, and I pick by intent. For a cold or mid-funnel asset I favour an in-player turnstile that appears after a real value hook, because it captures the person while they are engaged rather than at the door. For a webinar or a heavier resource, a short landing-page form before playback is fine, since the person arrived already wanting the content. When I am unsure, I run both against the same traffic for two weeks and let completion rate decide.
Not the view count. I track watch-to-opportunity: of the people who cross 50 to 75 percent of a key demo, how many become routed leads and then real opportunities. That number ties the video to pipeline instead of vanity, and it is the one I report to a client. View counts and even play rates can climb while none of it reaches sales, so I treat them as diagnostics, not results.
The core idea can be shared, but the opening cannot. Paid placements need the hook and the brand in the first two seconds because you are interrupting someone, and a stopping frame in the thumbnail earns the play. Organic and on-page video can breathe a little longer because the viewer chose to be there. I shoot once with a portrait-first eye and then version the cut per placement rather than producing two separate shoots, which keeps the cost sane.
