Last week I showed you that I created an AI clone that's published 60 videos in 90 days.

The number one question I got was:

“Mimmo, how do you actually do this?”

So today I’m showing you the exact system.

Step by step.
No fluff.
No tech jargon.

I’m not a software developer. I’m not a tech guy. I’m a cleaning business operator running 6 companies across Melbourne who hates being on camera.

If I can do this, anyone can.

By the end of this article, you’ll know exactly how to:

• Record content in 10 minutes without filming

• Turn that into professional videos using AI

• Publish 8 videos per week while running your business

• Do it all for about $50 per video instead of $3,800

I’ll show you the tools.
The mistakes I made.
And why this system works specifically for flat-out, camera-shy service business owners.

Let’s get into it.

The Complete System: 5 Simple Steps

If you can record audio on your phone and send an email, you can do this.

Here’s the entire system:

Step 1 — Record Your Training Footage (Once Only)

A 15-minute Zoom call where you train the AI on your face, voice, and speech patterns. You do this once. Never again.

Step 2 — Record Audio for Your Video

5–10 minutes talking into your phone like you’re explaining something to a customer.

Step 3 — Upload to HeyGen

The AI turns your audio into a video using your avatar. Takes about 20 minutes to generate.

Step 4 — Edit and Polish

Add B-roll, graphics, overlays, captions. About 3 hours using templates.

Step 5 — Publish Everywhere

YouTube, Instagram, LinkedIn, TikTok, Facebook, website.

That’s it.

Now let’s break each step down properly.

Step 1: Creating Your Training Footage (Do This Once)

This is the only time you need to be on camera.

15 minutes on Zoom.

You’re teaching the AI:

• What you look like

• How you sound

• How you speak naturally

Setup Requirements

Lighting

Face a window or turn on lights.
No studio setup required — just make sure your face is clearly visible.

Background

Keep it simple. Plain wall or office.

Audio

Most important factor.
Use headphones with a mic or sit close to your laptop mic.

Camera Position

Eye level. Not looking down or up.

Clothing

Wear what you normally wear in business.
Your avatar will always wear this outfit.

The Critical Rule

Speak naturally.

Don’t script.
Don’t act professional.
Don’t overthink it.

Talk like you’re explaining something to a customer.

The AI learns your real communication style.

I recorded mine in one take.
Uploaded it.
Done.

You never film again.

Step 2: Recording Audio for Each Video

This is where the system becomes powerful.

I record in my van between jobs.

Engine off.
Voice recorder app open.
10 minutes of talking.

That’s it.

My Process

I open Voice Memos on my iPhone.
Hit record.
Talk.

I don’t script word-for-word.

I use bullet points.

For example, for carpet pricing:

• Why prices vary by suburb

• What affects cost

• Why cheap quotes are a red flag

• What customers should actually pay

Then I speak naturally.

Yes, I say “um.”
Yes, I repeat myself.
That’s fine.

The clone mirrors your real voice.

Audio Quality Rules

• Quiet space

• Phone ~20cm from mouth

• Speak normally

• Don’t aim for perfection

If you mess up, pause 2 seconds and restart the sentence.

Jane edits it later.

Time Commitment

• 10 minutes recording

• 30 seconds sending file

Done.

Back to running businesses.

Step 3: HeyGen — Turning Audio Into Video

This is where audio becomes video.

I use HeyGen.
Other tools exist, but HeyGen handles Australian accents best in my experience.

First-Time Setup

Upload your 15-minute training footage.

Processing time: 2–4 hours.
You only do this once.

Creating a Video

1. Click “Create Video”

2. Select your avatar

3. Upload your audio file

4. Click generate

Generation time: 20–30 minutes for a 10-minute video.

You walk away.
It processes.

Download the finished 1080p MP4.

Editing Phase

Jane handles this part.

She adds:

• B-roll from real jobs

• Text overlays

• Graphics

• Subtle background music

• Intro/outro

• Captions

Editing time: ~3 hours using templates.

Without templates: 6–8 hours.

Cost Breakdown

HeyGen Business Plan:
$149 USD/month (~$230 AUD)
Unlimited video creation.

If producing 32 videos/month:

≈ $7 per video (HeyGen cost)

With editing included:

≈ $50–$70 per video total.

Compared to:

$3,800 traditional video production.

Massive difference.

Step 4: Editing Properly

Editing makes the difference between boring and engaging.

Jane:

• Cuts dead air

• Tightens pacing

• Adds contextual B-roll

• Inserts text highlights

• Designs thumbnails

• Optimizes YouTube SEO

Templates make this sustainable.

Same:

• Intro

• Outro

• Fonts

• Style

• Structure

Efficiency matters.

Step 5: Publish Everywhere

One 10-minute video becomes:

YouTube

Full-length version.

Instagram

5 short Reels cut from main video.

TikTok

Same short clips with adjusted overlays.

LinkedIn

Native upload for reach.

Facebook

Native upload.

Website

Embedded in blogs and service pages.

Email

Sent to mailing list.

Content Multiplication

One 10-minute recording becomes:

• 1 YouTube video

• 5 Reels

• 5 TikToks

• 1 LinkedIn video

• 1 Facebook video

• Website embed

• Email asset

Minimum: 14 pieces of content.

8 recordings per week = 100+ pieces of content weekly.

From less than 2 hours of recording time.

That’s leverage.

The 5 Mistakes I Made

Mistake 1: Scripting Word-for-Word

Result: Robotic delivery.

Fix: Bullet points only.

Mistake 2: Background Noise

Air conditioning ruined early videos.

Fix: Quiet environment.

Mistake 3: Perfectionism

Re-recorded endlessly.

Fix: Aim for helpful, not perfect.

Mistake 4: No Editing Templates

8-hour edits.

Fix: Standardized templates.

Mistake 5: Not Being Transparent

Initially hid AI usage.

Fix: Full transparency. Every video states AI assistance.

Customers care about value, not production method.

What Happens Next

Now you know:

• The 5-step AI clone system

• The tools (HeyGen, Voice Memos, CapCut)

• The cost ($50–$70 vs $3,800)

• The time commitment (10 minutes recording)

• The mistakes to avoid

You have two paths:

Option 1 — DIY

Set up HeyGen.
Record your training footage.
Start creating.

Completely doable.

Option 2 — Done For You

AFCB handles:

• AI clone setup

• Production

• Publishing

• Optimization

You just record 10-minute audio files.

Because here’s the reality:

Your competitors will figure this out.

Some already have.

You can be ahead — or behind.

I’d rather you be ahead.

I’m Mimmo.

See you in the next one.