Test different voiceovers for the same ad

The voiceover is a creative variable that marketers miss. It directly shapes the pacing, credibility, and emotional register of your entire asset, but you’ll often overlook it.

A calm, measured voice might be exactly right for a premium brand. An energetic delivery might be what closes the deal on a promo offer. A “founder voice”, conversational, slightly unpolished can do heavy lifting for a product that requires real trust.

The core advantage of TTS is that we can test all of these without booking a single studio session.

Instead of chasing the “perfect” voiceover out of the gate, treat the voice as a variable in an A/B test. Keep the visuals, the editing, and the copy locked. Change only the voice.

This isolates the variable and makes it much easier to diagnose whether a weak ad is a message problem, a pacing problem, or simply a delivery problem.

Do it now:

• Pick one existing ad: Pull an ad you’ve already built using AI.
• Use our prompt as your baseline: Generate your test versions from this script:

Every night, your body tries to recover. But stress, phone screens, and an irregular daily rhythm can make sleep feel less like rest… and more like a reset that never fully happens. Meet NOVA Sleep — a smart sleep band that analyzes your sleep, helps you wind down before bed, and wakes you at the best moment of light sleep [read more…]

• Generate several versions: Produce a few audio takes in completely different styles: tone, energy, pace, and register.
• Lock everything else: Keep the visuals, CTA, and ad length identical across every version.
• Run them as separate ads: Each voice variation gets its own placement.
• Measure what matters: Compare CTR, view rate, watch time, and conversions across the board.
• Lock in the winner: Save the top-performing style as your definitive voice direction for future creatives.

The output: Option 1 | Option 2 | Option 3 | Option 4

The voice used in our test recording is Orus.

To put the scale of this opportunity in perspective: a 2025 report projects the AI voice generator market will grow from $4.16B in 2025 to $20.71B in 2031. A 30.7% CAGR.

Turn your newsletter into a short audio briefing

Who says you need to read the newsletter. Especially when the content is dense and tactical.

Instead of asking your readers to scan a wall of text, fire up your TTS tool and convert your latest edition into a tight, 2–3 minute audio brief.

Think of it as a hybrid between a mini-podcast and a voice note from an expert friend who skipped the small talk. It’s a surprisingly underutilized format.

Your readers can catch it in the car, on a walk, or between meetings. And in the process, you’re building a more personal, human connection with your brand than a block of text ever could.

Step by step:

• Take your finalized newsletter draft: Start with the version that’s ready to send.
• Condense it to 350–450 words: Ruthless editing is the job here.
• Keep only what matters: The most important topics and core examples stay. Everything else goes.
• Add a brief intro and outro: Make them recognizable and consistent across every edition.
• Generate the audio with a consistent voice style: This is how you build a sonic brand identity over time.
• Embed the audio at the top of your email: Link or file, either works. Put it where readers see it first.
• Repurpose the audio: Drop that same track as a background for a Short, a Reel, or a LinkedIn post.

Here is an example of our prompt:

[calm intro] Welcome to AI Radar Brief — your quick audio rundown of the most interesting signals from the world of artificial intelligence.[short pause] [confident] Today: AI-generated worlds, self-replicating agents, DeepMind’s math assistant, and AI helping astronomers discover new planets.[read more…]

Listen to the output.

Pro tip: By selecting the “energetic co-host podcast style conversation” prompt or setting in your AI audio tool, you can generate a dynamic dialogue between two characters.

It’s a frictionless way to launch your own automated podcast without producing a single episode from scratch.

Create a brand song using Lyria 3

AI audio doesn’t stop at voiceovers. One of the most interesting emerging use cases is generating short songs, jingles, and sonic brand motifs.

Tools like Google’s Lyria 3 let you rapidly test musical directions during the early concept phase.

And to be clear: you don’t need to aim for the Billboard Hot 100 here. The goal is to develop a specific, repeatable sound that people instantly associate with your business.

For Tactics, we could create a short track that reinforces the newsletter’s positioning: practical AI workflows, less hype, more action, and a competitive edge for marketers.

That snippet can act as an intro for an audio briefing, background music for Shorts, a video outro, or a core element of a promo campaign.

The key is that the song isn’t random. It has to sell the brand’s personality.

The Playbook:

• Define exactly what your brand should sound and feel like.
• Pick a few core phrases that absolutely must appear in the lyrics.
• Choose a musical genre that resonates with your target audience.
• Generate a few different versions of the track.
• Cut the best 10–15 seconds to use as a standalone jingle.
• Test it across your Shorts, ads, and audio briefings.
• Save the winning track as a permanent piece of your sonic branding.

Here is an example of our prompt:

Create a modern, catchy brand song for an AI newsletter called “Tactics”. Brand context: Tactics is a weekly newsletter about the newest AI tools and practical AI strategies. It is created mainly for marketers, founders, entrepreneurs, business owners, and people managing companies. [read more…]

And you can listen to the banger output here.

Pro tip: You can use video generation models like VEO 3.1 or Seedance 2 to create a music video using your new song, which strengthens the message.

The sounds that matter

If there’s one thing these tactics have in common, it’s this: audio is your strategic lever that most of your competitors still treat like an afterthought. Here’s the short version:

• Treat voice like a creative asset: Test three different TTS styles on the exact same video and measure the hard impact on your CTR and conversions.
• Turn your newsletter into an audio brief: Win back busy readers by giving them a punchy 3-minute track to play in the car instead of a wall of text to scan at their desk.
• Build your sonic DNA: Generate a short jingle with Lyria 3 to tie your ads and Shorts together — no composer budget required.

Happy generating.