How do you make a video ad without appearing on camera?

Six approaches work: an AI presenter, a screen recording, product footage, hands-only shots, voiceover over b-roll, or text on screen. Vidpop takes the first route — its presenter is an AI avatar, so no real person is filmed — and pairs it with either your app screen recording or an animated clip made from your product photos.

Do faceless video ads perform worse than ones with a person?

It depends far more on the script than on the format, in our view — a specific, well-written faceless ad routinely outperforms a vague one with a person in it. What a face reliably adds is warmth and direct address, which is why tools like Vidpop use an AI presenter rather than dropping the presenter entirely.

Can Vidpop make an ad with no presenter at all?

No. Every Vidpop reel includes an AI presenter and there is no presenter-free mode, so it solves the problem of not wanting your own face on screen but not the problem of wanting no face at all. For a completely person-free ad you need a different tool or an edit pass elsewhere.

Do AI presenters look fake?

Some viewers notice and some do not, and no honest answer is more precise than that. Vidpop uses stock AI avatars — 10 in the English catalogue and 4 in the Arabic one — which means you can pick a presenter that suits the audience, but you cannot train a custom avatar of yourself or change their clothing.

What is the cheapest way to make a video ad with no face in it?

Text on screen over your own footage, with a voiceover, costs nothing but time and works well in muted feeds. The trade is that all the persuasion lands on the writing, which is also true of the faceless portion of any ad — including the demo section of a Vidpop reel, where the presenter shrinks into a corner while the product fills the frame.

Can I use my product photos instead of filming?

Yes, and animating them beats showing them static. Vidpop's product flow takes one to four photos, generates a hero still from them, and animates it into an 8-second demo clip, which is enough movement to read as video rather than as a slideshow.

Is it better to hire a creator than to use an AI avatar?

If you have a physical product you can ship and a real budget, a human creator holding your actual product carries authenticity an avatar cannot fake — Billo's marketplace exists for exactly that. If you sell software, work in minutes rather than weeks, or have no product to post, an AI presenter in something like Vidpop is the practical option.

Do faceless ads still need captions?

More than any other format. With no face to read and often no sound playing, captions frequently carry the entire message. Vidpop burns captions into every reel and generates their text from a transcript of the rendered voiceover, so the words on screen match what is actually said.

Guide

Making video ads without showing your face

Updated 28 August 2026

Not wanting to be on camera is an ordinary position and has very little to do with confidence. Plenty of people simply do not want their face attached to a product feed, or in an ad that will still be running in two years. There are six workable approaches. Each of them trades away something real, and the pages that list them rarely say what. This one does.

The six approaches, side by side

	Best for	Where it falls down
AI presenter	Keeping a face and eye contact on screen without it being yours	Reads as synthetic to some viewers, and you cannot control what they think of that
Screen recording	Apps and software, where the product is the screen	Nothing to show if what you sell is physical
Product footage	Physical goods you can photograph or film	Static product shots look flat without motion or a person
Hands-only shots	Anything held, used, unboxed or assembled	It is a real shoot — camera, light, framing, retakes
Voiceover over b-roll	Fast turnaround with no shoot at all	Generic stock footage reads as generic, and viewers notice
Text on screen only	Feeds watched with the sound off	Every bit of the weight lands on the writing
A hired creator	Real human credibility with someone else's face	Costs real money and usually needs a product you can ship

What each one actually costs you

An AI presenter is a synthetic person delivering your script to camera. It is the only approach here that preserves what a face does in an ad — direct address, eye contact, someone appearing to talk to you — without anyone sitting in front of a lens. The trade-off is worth stating plainly: some viewers clock it as synthetic, and some of those hold it against the product. No amount of avatar quality removes that entirely.

A screen recording is unbeatable for software, because for software the screen is the product. It costs nothing but your time, it is inherently truthful, and it shows rather than tells. It is also useless if you sell a chair.

Product footage covers everything from a proper shoot to photographs animated into motion. Photos alone look like a catalog; movement is what makes a still feel like video. If you already have good product photography, this is the cheapest route to something deliberate.

Hands-only shots are the sleeper option. A pair of hands using a product carries a surprising amount of the credibility a face would, with none of the identifiability. The catch is that it is an actual shoot — camera, stable surface, decent light, and the patience for five takes of picking something up.

Voiceover over b-roll is the fastest thing here and the easiest to do badly. Stock footage of someone smiling at a laptop signals "advertisement" faster than almost anything else in a feed. It works when the b-roll is specific to you and falls apart when it is not.

Text on screen only is the purest form, and genuinely effective in feeds where most people watch muted. But there is nowhere to hide: with no voice and no face, mediocre copy that would survive a warm delivery dies on a title card.

Step by step: making one

01Pick the approach from what you sell, not from what you like. Software points at screen recording. A physical product points at product footage or hands. A service with nothing to show points at an AI presenter or strong text.
02Write the script first, before you gather a single visual. Thirty seconds is around 70 to 80 spoken words. Structure it as hook, demonstration, call to action, and decide the one problem you are naming before anything visual.
03Gather the visuals the script actually needs. Read each line and ask what is on screen while it is spoken. Lines with no answer are lines to cut.
04Decide who is speaking. A voice with no face still has a personality, an age and an accent, and that choice does as much work as any visual here.
05Record or generate the voiceover at the pace of the footage. A read that races ahead of what is on screen feels anxious.
06Add captions. Not optional. Much of feed video is watched muted, and in a faceless ad the captions are frequently the only thing carrying meaning.
07Put a real call to action at the end, on screen and in the audio. With no presenter to ask, the closing card doing the asking matters more, not less.
08Watch it once with the sound off, then once with your eyes off. If it works in both, it will work in a feed. If it only works with both, it is too fragile.

Where Vidpop fits, and where it does not

Vidpop is an iPhone app that makes short vertical video ads, and its answer to this problem is the first approach on the list: the presenter is an AI avatar, so nobody involved has to be on camera. There are 10 presenters in the English catalogue and 4 in the Arabic one, and the voice is decoupled from the face — each presenter has a default voice, but you can change either one independently, choosing from 10 English voices and 7 regionally specific Arabic ones.

It fits the two visual approaches that pair with a presenter. For an app, you supply your own screen recording and Vidpop composites it inside a drawn iPhone frame over a generated background, shrinking the presenter into a picture-in-picture during the demo instead of cutting away. For a physical product, you give it one to four photos and it generates a hero still, then animates that into an 8-second demo clip with the presenter in a side picture-in-picture.

Now the part that matters if you are reading this page specifically. Every Vidpop reel contains an AI presenter, and there is no presenter-free mode. You cannot turn the avatar off and keep the screen recording, the voiceover and the captions. If you want an ad with no face in it at all — pure product footage, pure text, pure hands — Vidpop is not that tool. It solves "not my face" completely and "no face" not at all.

The other limits: presenters are stock AI avatars, so you cannot train a custom avatar of yourself or change a presenter's clothing or pose. Output is fixed at 720 × 1280, 9:16, about 25 to 35 seconds. English and Arabic only. iPhone-only, iOS 17 or later, no Android or web version, and a subscription is required to render.

If you want a human face, just not yours

There is a seventh option outside the DIY list, and it is the right answer often enough to name. Creator marketplaces put a real person in front of the camera on your behalf. Billo is a human creator marketplace rather than an AI generator — their own site says "Human creators, not AI avatars" — where you brief a campaign, a vetted creator receives your physical product, films a real review video, and delivers it with full IP transfer to the brand.

That is a different budget and timeline from generating something on your phone, and it needs a physical product you can ship. But if the reason you want a faceless ad is that you do not want to be in it — rather than wanting no human in it — a real person holding your real product carries an authenticity no avatar structurally can.

What goes wrong

It feels cold. Faceless ads lose the warmth a person supplies, so the script has to work harder. Write more directly and more specifically than you would for a talking head.
Viewers comment that it is AI. They will sometimes. The mitigations are a script that sounds like a human wrote it, a voice that suits the brand rather than the default, and keeping the ad short.
The voice does not fit the face. In tools where the two are locked together you are stuck. In Vidpop they are independent, so swap the voice rather than accepting the pairing you were given.
There is nothing to film. If the product is a service or an idea, lean on text, motion and a presenter instead of stock clips of strangers in offices.
The b-roll is doing nothing. If a clip could be swapped for any other clip without changing the meaning, it is filler. Cut it and let the text hold the frame.
Nobody knows what to do at the end. With no presenter to ask, the closing card is the entire ask. Make it a verb and a destination, not a logo.

Frequently asked questions

How do you make a video ad without appearing on camera?: Six approaches work: an AI presenter, a screen recording, product footage, hands-only shots, voiceover over b-roll, or text on screen. Vidpop takes the first route — its presenter is an AI avatar, so no real person is filmed — and pairs it with either your app screen recording or an animated clip made from your product photos.
Do faceless video ads perform worse than ones with a person?: It depends far more on the script than on the format, in our view — a specific, well-written faceless ad routinely outperforms a vague one with a person in it. What a face reliably adds is warmth and direct address, which is why tools like Vidpop use an AI presenter rather than dropping the presenter entirely.
Can Vidpop make an ad with no presenter at all?: No. Every Vidpop reel includes an AI presenter and there is no presenter-free mode, so it solves the problem of not wanting your own face on screen but not the problem of wanting no face at all. For a completely person-free ad you need a different tool or an edit pass elsewhere.
Do AI presenters look fake?: Some viewers notice and some do not, and no honest answer is more precise than that. Vidpop uses stock AI avatars — 10 in the English catalogue and 4 in the Arabic one — which means you can pick a presenter that suits the audience, but you cannot train a custom avatar of yourself or change their clothing.
What is the cheapest way to make a video ad with no face in it?: Text on screen over your own footage, with a voiceover, costs nothing but time and works well in muted feeds. The trade is that all the persuasion lands on the writing, which is also true of the faceless portion of any ad — including the demo section of a Vidpop reel, where the presenter shrinks into a corner while the product fills the frame.
Can I use my product photos instead of filming?: Yes, and animating them beats showing them static. Vidpop's product flow takes one to four photos, generates a hero still from them, and animates it into an 8-second demo clip, which is enough movement to read as video rather than as a slideshow.
Is it better to hire a creator than to use an AI avatar?: If you have a physical product you can ship and a real budget, a human creator holding your actual product carries authenticity an avatar cannot fake — Billo's marketplace exists for exactly that. If you sell software, work in minutes rather than weeks, or have no product to post, an AI presenter in something like Vidpop is the practical option.
Do faceless ads still need captions?: More than any other format. With no face to read and often no sound playing, captions frequently carry the entire message. Vidpop burns captions into every reel and generates their text from a transcript of the rendered voiceover, so the words on screen match what is actually said.

Try it on your own app or product

Free to download on iPhone. A subscription is needed to render a reel.