How to Make Videos Without Showing Your Face

The obstacle to posting video without your face is rarely the idea. It is production. Filming, lighting, retakes and editing eat the evening, and most tools that promise “easy” still expect you to assemble everything by hand.

This is the production side: what to do once you know what you want to make, from a written script to a file you can post. If you are still deciding on a format, start with the ten faceless video formats and what each one needs, then come back here.

The whole loop takes under ten minutes per video once you have done it twice.

Step 1: Write to length, not to feel

Speaking pace is roughly 2.5 words per second, so length is arithmetic, not instinct:

• 15 seconds: 35 to 40 words
• 30 seconds: 70 to 80 words
• 45 seconds: 110 to 120 words
• 60 seconds: 145 to 155 words

Write past the target, then cut back to it. A first draft is usually about 30% longer than the version that holds attention. For the shape of the script itself, use Hook, Setup, Value, CTA.

One idea per video. If you have three, you have three videos.

Step 2: Choose the voice

Three options, with real tradeoffs:

Your own voice. Highest trust, no cost, and it stays consistent for free. The catch is that it ties every video to a recording session, which is exactly the friction you were trying to remove. It also is not faceless in the way some people need: a voice identifies you.

Text to speech. Fastest, fully batchable, and no recording setup. Modern voices are good enough that most viewers do not notice. Pick one voice and keep it across every video, because switching voices between posts reads as a different channel.

A voice you never speak in. If privacy is the actual reason you are here, text to speech is the only option that holds. Recording your own voice under a pseudonym works until someone who knows you scrolls past.

Whichever you choose, read the script out loud before committing. If a sentence makes you stumble, it will make a synthetic voice stumble too.

Step 3: Decide what is on screen

The script determines this more than you would expect.

A character works when the script is explanation or opinion, anything where a person would normally be talking. It solves the recognition problem faceless channels have: viewers remember a character the way they remember a presenter. Build it once, reuse it in every video, and do not redesign it.

Stock footage works for mood, narrative and anything abstract. It fails when the clips feel generic, which happens fast if you take the first result for every search.

Screen recording works for anything that lives on a screen. Nothing beats it for tutorials.

Text alone works when the writing is strong enough to carry the video by itself, and it plays fine on mute.

Mixing two is normal. Mixing four is a mess.

Step 4: Captions are not optional

Most of the feed is watched on mute, and a video with no captions is a video with no audio track for most of its viewers. Burn them in rather than relying on platform auto-captions, which are inconsistent and are not always shown.

Keep them large, high contrast, and in the middle band of the frame. Two to four words per line reads faster than full sentences.

Step 5: Export for the platform, not once for all of them

TikTok, Reels, Shorts: 9:16, 1080x1920. Keep anything important out of the top 15% and bottom 20% of the frame, because the interface covers it.

LinkedIn and feed posts: 1:1 or 4:5 uses more screen in the feed than 16:9.

YouTube long form: 16:9, and the first three seconds decide the rest.

Pick the cover frame deliberately. On grid views it is doing the work of a thumbnail.

Step 6: Batch, do not drip

Write five scripts in one 45 minute session with the timer running and no editing. Produce them in one sitting after that. Two focused sessions per week outproduce daily improvisation, and consistency is what the feeds reward.

Keep a running list of hooks and topics somewhere plain. The empty page is what kills faceless channels, not the production.

Where this usually goes wrong

The first line. “Hey guys, today we’re going to talk about” is where most faceless videos lose. Start inside the content: “In 1998 a bank paid $50 million for a company that did not exist.”

Redesigning the character. Recognition compounds only if the thing stays the same.

Adding a second idea. It halves the completion rate and completion rate is what gets you distribution.

Waiting for better gear. There is no gear in this workflow. That was the point.

The short version

Write to a word count, pick one voice and keep it, put one kind of thing on screen, burn in captions, export vertical, batch the writing.

MonkeyTalker runs steps 2 through 5 from a pasted script: the character lip-syncs the narration, captions and b-roll are generated, and the export is already sized for vertical. It takes about three minutes per video once the character exists, and the free plan needs no card. Write one 75-word script today and put it through, so the next one is faster.