Day 14: Cracking the Video Sales Letter Challenge With AI

Yesterday’s results were modest: $561 in sales against £330 ($447) in ad spend, leaving just over $120 profit. Not a big win, not a disaster. Just one of those days where the graph ticks along and you keep going.

But today’s story isn’t about the numbers.

It’s about a creative challenge I’ve wrestled with for months: could adding a video sales letter (VSL) to my funnel increase conversions?

The Dilemma of “Being on Camera”

My sales page so far has been text-only. It works, but I’ve always wondered if a video could lift conversions from 2% closer to 3% — a small percentage change that could transform the entire funnel.

The problem?

• I’m not the guitar teacher. I can market the product, but I’m not the face of the lessons.

• I don’t feel comfortable on camera. The thought of presenting a course I didn’t create makes me feel like an imposter.

For weeks, this held me back. But then I asked myself: what if I didn’t need to be on camera at all?

Experimenting with AI Video Tools

That’s when I discovered tools like HeyGen, which can create digital avatars.

I uploaded a photo of myself and watched in awe (and mild horror) as a clone of me, complete with an oddly Australian accent, delivered a sales script.

It was surreal — and not quite right. But it opened a door.

Instead of trying to fake a perfect on-camera performance, I decided to build a voice-led, image-driven video:

• I wrote a short four-minute script about the offer.

• I asked AI to break it into 10 key visual beats and generate prompts for images.

• I recorded the narration in my own voice, synced to those images.

• The software let me tweak words in the script afterwards — without re-recording.

The result?

A “rough and ready” VSL that’s not slick, but good enough to test.

From Rough Draft to Split Test

This first version won’t win any awards. Some images look a little “AI-weird” (one guitar looked like it belonged in a Salvador Dalí painting).

But that’s fine.

The goal isn’t perfection. The goal is proof of concept:

• Does a video version of the sales page convert better than the text-only page?

• If so, I’ll invest more time in refining it.

• If not, I’ll know not to waste weeks polishing a strategy that doesn’t move the needle.

Today, I’ll launch a split test: one page with video, one without. Within days, I’ll know if the VSL has legs.

“Small percentage shifts at the top of the funnel — like moving from 2% to 3% conversions — can transform the whole business.”

What’s Next

Tomorrow marks the two-week milestone of this journey. I’ll be pulling together the numbers from the first 14 days:

• Where we’ve made progress.

• What’s failed (plenty).

• And the lessons that will guide the next phase.

But for now, I’m excited. Because whether this test wins or loses, it’s another step forward in building a funnel that works.