Only a handful of the people who hit your page ever press play. Here is what moves the two numbers that decide whether your VSL works, and what to do about the apply button
Most of them never press play.
5 acts14 sections16 min read
ACT I
Read This First
Before the arguments about buttons, thumbnails and length, get the scoreboard straight.
THE SCOREBOARD
Two numbers decide whether a VSL works
Everything in this playbook moves one of these two numbers.
1. Play rate
Of everyone who hits your page, how many actually press play. Before anything else matters, this number has to happen.
2. Engagement rate
Of the people who pressed play, how much of the video the average person actually consumes.
So we are talking about the same thing.
If you are running a VSL call funnel right now, spending real money on ads, and your cost per qualified call is climbing while your sales team says nobody watches the video, this is for you. If you have not built the funnel yet, read it anyway, because most of the damage happens in the setup.
Two numbers decide whether a VSL works. Almost nothing else you argue about does.
ACT II
The First Ninety Seconds
How to structure the video so the people in a hurry still get the point.
STRUCTURE
Say everything in the first 60 to 90 seconds
Treat the first 60 to 90 seconds like it has to contain everything the whole VSL covers, said in that tiny space. Then spend the rest of the video expanding on each point.
Everything, compressed. Then expand it.
Most people open a VSL with something they consider a hook, or they dip straight in. One high-ticket program's VSL currently opens by saying the price, straight away, then covers everything people want to know inside the first 60 to 90 seconds.
Here is why it works, and it is not flattering to anyone. When people land on your page, they want efficiency and energy conservation. They want to decide whether this is for them, and whether to act, with as little time invested as possible. They do not care about watching your whole video the way you care about them watching it. Those are competing incentives.
Think the tweet version of what you would otherwise write as a newsletter. Then expand.
They are not deciding whether to keep watching. They are deciding whether to leave.
CHAPTERS
Chapter the video
Add chapters to the player. Every video player supports it now. When someone wants to skip around, they hover and scrub straight to the section they care about, because nobody wants to watch it all in the order you built it.
Example chapters. Name yours after what people actually want to know.
Many more people will go through the information the way they prefer to. Some of you already skipped the last section and never read this.
Chaptering helps retention and consumption. When someone is done with what you are saying, they move on anyway. The chapters just decide whether they move on inside your video or off your page.
They are going to skip. Decide where to.
ACT III
Should You Withhold The Apply Button?
The argument everybody has. What actually decides it is your bank account.
THE DECISION
What actually decides it is your bank account
Does the apply button show the moment somebody hits the page, or only after they are a certain way into the VSL?
Here is how most people arrive at it. Leads come through, a lot of them have not watched the video, and you get frustrated. So before changing anything else, you put a time delay on the button. Some test a minute or two. Others with very long VSLs hold the button until twenty or thirty minutes on the page.
Withholding the button dramatically reduces how many people schedule and dramatically inflates your cost per qualified call, at least in the short term, while the pixel builds new data on people more likely to watch and then act. Sometimes you see an immediate benefit. A lot of the time you do not.
The real question is whether you can afford to sit in that period.
THE PIXEL
You are also starving the pixel
There is a second cost, and it does the long-term damage. When you withhold the button, fewer people convert, so fewer conversions get sent back to the pixel. That breaks pixel-conditioning best practices.
Right now, in 2026, getting out of learning mode matters. In some periods it has not mattered at all, and some campaigns sitting in “learning limited” still rip. But right now, exiting the learning phase makes a big difference to performance.
The upside, if you have enough scale to get the data back: the right people consume a lot of the VSL and then apply, which raises intent. You follow pixel-conditioning best practice and improve lead quality at the same time, because only high-intent leads land in the results column.
Direction, not a formula. The threshold moves account by account.
Fewer applications means fewer signals sent back.
FLIP IT
The button trade
Flip the button and watch all four consequences move at once. Three of them go against you.
Direction only, on purpose. “Dramatically” is the honest description; there is no fixed figure.
Withholding the button is not a tactic. It is a purchase, and the price is your cost per qualified call.
THE CONDITION
The setting depends on who takes the calls
Withholding can help if you can endure the process. It is not necessary. And it depends on your sales team.
A team of killers → show the button
They do not care whether people watched the full video. They will nurture and close them regardless. Withholding does not make much sense.
A team that needs warm leads → withhold, or switch funnels
If your team needs people to have consumed content before the call, you will have to withhold. You might even switch to a webinar or a content-first strategy.
And there is the other direction: a sales team demanding you fill their calendars with as much volume as you can get, qualified or not. Whichever is actually your situation, lean into it.
The button setting is a sales-team decision wearing a marketing costume.
THE PLAY BAR
If you hide the button, hide the play bar
Showing the video's play bar is a separate argument. Here is what holds up:
Setup
What happens
Bar shown + chapters, button visible
WorksWhat most people run, with a lot of benefit.
Bar hidden, button delayed
WorksConsistent: nobody can scrub past your time threshold.
Bar shown, button delayed
BreaksSomeone scrubs the whole video in a fraction of the time, feels ready, and there is nothing to click.
Bar hidden, button visible
BreaksFeels smart, but lowers intent. People who want to skip ahead are forced to sit through it, so you attract lower-quality leads.
Hide both or hide neither. Hiding just the bar buys you worse leads.
ACT IV
Getting Them To Press Play, And Stay
The two numbers, one at a time.
PLAY RATE
The thumbnail is the ad for the ad
Proven again and again through 2025 and into 2026: video thumbnails still absolutely rip.
The thumbnail should show the thing people are most likely to care about when they hit the page. What would they want to know? Answer it in the 5 to 10 seconds you get, and it invites them in. Pick a section of the video, or shoot a separate short clip just to be the thumbnail.
A still image is also justifiable, because a video thumbnail slows the page down a little. The concept is identical: it has to show what the video covers, and what you flex has to be something the viewer actually cares about.
Video player
Verdict
Wistia
UseOne of the two actually being used
Vidalytics
UseOne of the two actually being used
YouTube, Vimeo, others
AvoidStay away from them
A few seconds to prove the video is worth it.
THE PAGE
Put less on the page
Put too much information on the page and people bias toward efficiency: they scroll it, decide whether it was for them, and leave or act. You have just negated the video. The layout that still gets run on VSL call funnels is three things:
your-funnel.com
Headline
▶ VSL
Application (or the button)
That is it. Isolate the assets so anyone who wants the information knows they have to watch to get it. A VSL is a curated presentation: information in a deliberate flow, with calls to action and proof that validates your claims and gives certainty. Watching it is a high-effort ask, so there has to be an incentive.
Headline, video, application. Nothing else.
DRAG IT
A high engagement rate can be a bad sign
A high play rate is good. A high engagement rate can be bad. It means one of two things: people are hooked and want everything you are saying, or they are interested but struggling to understand what you do and how it works, so they keep watching to find out.
Very lowVery high
Technically you want engagement as high as possible. But if you can prompt the action and give people the certainty to take it, while keeping them qualified, getting your point across in fewer words is the better outcome.
ENGAGEMENT
Editing buys attention, congruency keeps it
People's brains are fried by how much competes for their attention in a day. Holding someone for the length of a VSL takes real work now.
Editing
Jump cuts and captions (you can get away with these alone), visual edits over the talking, and the occasional music, sound effect or meme moment. Retention improves with each.
Communication
Your tonality, the words you choose, messaging that stays congruent with the ad or video that sent them, the authority you speak with, and the storyline through the whole VSL.
The second column is not editing at all, and it is where a lot of people mess up.
Editing buys you attention. Congruency keeps it from being wasted.
ACT V
Length And Style
How long, and in what form.
LENGTH
Length is an audience decision
The shorter and more direct you are, and the more concentrated the information, the better you usually do with a wealthier demographic. The general public responds well to longer content; one business right now runs a literal podcast as its VSL, unedited, intro and all.
Length is about who is watching.
Rich people prefer shorter VSLs, and that genuinely decides it. If you run a longer VSL at a rich audience, use the trick from Act II: compress everything into the opening 60 to 90 seconds, then elaborate, so they can decide the moment they feel certain enough.
Who you sell to changes the asset.
STYLE
Three styles, an even mix
Three styles are working, and it is honestly an even mix. None dominates.
Style
What it looks like
Google Doc read-along
Screen shared, reading through the doc
Classic talking head
Has worked for a long time and is still very effective
Slides with a voiceover
Basically a screen recording. Good if you would rather not be on camera. Tried and true.
Whichever you pick, ten to twenty minutes is usually the norm. There are people going well beyond, and there are outliers like the podcast example.
There is no winning style. There is the one you will keep making.
THE PLAYBOOK LIBRARY
Keep going: fix the next thing
9 free playbooks on the problems we fix for clients.