How to write a 90-second homepage video script that actually sounds like you

Most homepage videos fail in the first ten seconds. Not because the lighting is bad or the audio is muffled — those are fixable. They fail because the person on screen is reading a script that was written to impress, not to connect. They open with "Hi, I'm Sarah, and welcome to my website" in the same cadence a YouTuber uses to ask for subscriptions, and the visitor — who came to find out whether to trust this person with their money — clicks away in the time it takes to say "today we're going to talk about."

A good homepage video script does one job: it makes the visitor feel they have already met you. Ninety seconds is enough. Less is fine. More is rarely better. What follows is a framework I have refined over years of helping small service businesses figure out what to actually say on camera.

Why 90 seconds is the right ceiling

Anything longer assumes the visitor has decided to invest in you, which they have not. Anything much shorter — your standard 30 second intro video — works for ad placements but rarely gives you room to do the real work, which is showing the visitor what kind of person you are.

Ninety seconds gives you space for five distinct beats: a hook, a moment of recognition, a differentiator, casual proof, and an invitation. Each beat has a time budget. The budgets matter because rambling is what kills these videos, not under-preparation.

I have watched dozens of solo founders and small agencies record this. The ones who hit the timing tend to feel natural. The ones who treat it as a monologue tend to sound like they are reading a brochure aloud.

The five-part script formula

Here is the structure. Treat it as a frame to write into, not a template to fill in.

1. Hook (0–15 seconds) — who you are, who you serve

Open with your name and the single sentence that names your audience. Not your job title. Not your credentials. The audience.

"I'm Riin, and I help small e-commerce stores look like they belong online."

That is fifteen seconds of work, but it does an enormous amount. The visitor knows within one breath whether the page is for them. If they are running a small Shopify store, they lean in. If they are not, they leave — which is a feature, not a bug. You do not want everyone. You want the right people.

What you should not do here is the YouTuber opener. "Hey guys, welcome back to the channel" is wrong for service businesses for the same reason a dentist greeting you with "what's up gamers" would be wrong. The register does not match the trust level you are trying to build.

2. The pain you understand (15–35 seconds)

Now you name one specific frustration your buyer has. Not three. One. The specificity is what makes them feel seen.

"Most of you have a Shopify store that looks like everyone else's Shopify store. The theme is fine, the product photos are fine, and somehow nothing converts the way you thought it would."

The work here is restraint. You are not listing every problem you solve. You are picking the one that, when said out loud, makes your buyer think yes, that's me. If you are a coach, it might be the loop of starting and stopping. If you are an accountant, it might be the dread of opening the email from the tax office. If you are an agency, it might be the fifth time a client has asked why their last designer ghosted them.

For audience-specific examples of where this shows up, the niche pages on Heyly for coaches, Heyly for agencies and Heyly for ecommerce each carry the kind of pain language that tends to land on a homepage video.

3. What you do differently (35–65 seconds)

This is the longest beat — about thirty seconds — and the trap is filling it with a list. Resist. Pick one concrete thing that distinguishes how you work and describe it like you are explaining it to a friend at a dinner party.

Not: "We offer a comprehensive five-step methodology covering strategy, design, development, launch, and optimisation."

Closer to: "Before I write a single line of code, I spend a morning on your existing site pretending to be a customer who has never heard of you. I write down every place I get confused. That document is what we work from."

The second version is memorable because it is specific. The first version is forgettable because it could have been said by anyone. Generic descriptions of process are exactly what the visitor expects, and meeting expectations is not how trust gets built. A small surprising detail — I spend a morning pretending to be your customer — is what they will remember when they tell their business partner about you tomorrow.

4. Proof, casually (65–80 seconds)

One specific result, said without bragging. Fifteen seconds is enough.

"Last project was a small jewelry brand — went from about 30 to 80 orders a month, with the same traffic."

The casualness matters as much as the number. A confident person mentions a result the way they mention what they had for lunch. They do not bold it, they do not capitalise it, they do not pause for effect. The result speaks. You just deliver it.

If you are early in your business and do not yet have a clean number, talk about a moment instead. "The last client I worked with sent me a screenshot of their first inquiry from the new site within three days of launch. That's the kind of thing I'm trying to make happen." That works too.

I have written separately about why testimonials, case studies and logos do different jobs — and the same logic applies to the proof beat in your video. You are not building a case study. You are dropping one line of evidence.

5. The ask (80–90 seconds)

Friendly, low-pressure, ten seconds.

"If that sounds like what you need, the contact button is right below."

That is it. Do not pitch. Do not list packages. Do not say "book a discovery call to unlock your potential." The person has just spent eighty seconds with you. They know whether they want to talk. Your job is to make the next step feel obvious and easy, not urgent.

Three mistakes that quietly kill these videos

The YouTuber opener is one. The hard pitch at the end is another — anything that smells like a closing line from a webinar lands wrong on a service-business homepage. And the third is over-rehearsal. I have seen people record a service business video twenty-eight times until it sounds polished and lifeless. The version they recorded on the third take, the one with the small stumble where they laughed at themselves, was the better video. They just did not trust it.

Polish is not what builds trust. Presence is. A small flub where you correct yourself reads as human, which is exactly the signal you are trying to send.

A note on what this script is not

It is not a template you fill in. If you copy the example sentences word for word, the result will sound like a copy of someone else's homepage video, which is worse than no video at all. The framework is there to give you a shape. The words are yours.

Read your draft out loud before you record. If a sentence feels like something you would never say to a friend over coffee, rewrite it until it does. That is the entire test.

Record it on your phone. Don't spend a week rehearsing. The point is to be a person, not a performer.

Frequently Asked Questions

How long should a homepage video really be?

Ninety seconds is a working ceiling, not a rule. If your message lands cleanly in sixty, use sixty. The reason I push back on longer videos is that visitor attention drops sharply after the first minute on a homepage, and anything past two minutes is essentially watched only by people who already wanted to hire you. For most service businesses, the sweet spot is somewhere between 60 and 90 seconds. Shorter is acceptable; longer needs to earn its time. If you find your script running past 100 seconds, the usual culprit is the "what you do differently" section turning into a list. Cut it back to one concrete thing and the rest tends to fall into place.

Do I need professional equipment to record this?

No. A modern phone in good natural light produces video that is more than acceptable for a homepage. What actually matters is audio — if your viewer cannot hear you cleanly, they leave within seconds, regardless of how good the picture looks. A cheap lavalier microphone clipped to your shirt solves ninety percent of audio problems. Beyond that, find a quiet room, sit near a window during daytime, and record. Production value is not what builds trust on a service-business homepage. Slightly imperfect video with a real person in it consistently outperforms polished corporate video, because polish reads as distance and distance is the opposite of what you are selling.

What if I freeze up on camera?

Most people do, the first three takes. The fix is not more rehearsal — it is the opposite. Write your script, read it twice, then put it away and record the video as if you were explaining your work to a friend who just asked what you do. The takes that sound natural are almost always the ones where the speaker stopped trying to remember the script word for word and started just talking. Keep your phone propped up somewhere comfortable, look slightly above the lens (not directly at it, which can feel intense on a small screen), and accept that your fifth take will probably be the keeper. If you stumble, laugh, correct yourself, and keep going. The viewer will read it as warmth, not error.

Should the video autoplay on my homepage?

Generally no, with sound. Autoplaying audio on a homepage is one of the fastest ways to make a visitor close the tab. What works better is a video thumbnail with a clear play button, ideally a still frame where you are smiling or mid-gesture, with a short caption like "30 seconds — what I do and who I do it for." Visitors who want to watch will click. The ones who do not will skim the rest of the page, and that is fine. The video is one trust signal among several, not the whole pitch.