This website uses cookies

Read our Privacy policy and Terms of use for more information.

[ TACTICS ]  /  AI marketing tactics

Three audio AI tactics that give your ads emotion, timing, and a payoff

October 11, 2026 6 min read
Presented by 90 Percent
Three audio AI tactics that give your ads emotion, timing, and a payoff
Sponsored by 90 Percent90 Percent

$10k in 1986 is ~$244k today, and 90% of pros couldn't do better

Hollywood loves a Wolf of Wall Street. In real life, the sheep in the index fund keeps winning.

The pros aren't the heroes here. Over time, 90% of professional investors fail to beat the index after fees.

And the index has receipts. Since 1986, the FTSE All-World has averaged ~8.5% a year through:

  • The 1987 crash
  • The dot-com bubble
  • The 2008 financial crisis
  • A pandemic

That turned $10,000 from 1986 into roughly $244,000.

The hard part is staying in. When markets drop, an index fund feels abstract. It's a lot easier to hold on when you know what you actually own.

So every Tuesday, 90 Percent pulls one random company from the FTSE All-World's 4,200 and tells you its story in 5 minutes.

No stock tips. Just the world you already own.

Meet your first company here.

Audio AI outgrew the “glorified text-to-speech” label a while ago.

That’s where things start to get interesting.

With the latest models, you’re no longer limited to telling them what to say or play. Increasingly, you can tell them how to do it.

A voice can slide from calm to full panic inside a single scene. Music can be composed around a video that’s already been cut.

You can even hand one melody around like a relay baton, from piano to guitar to cello to brass.

Put together, that shifts audio from something you bolt onto a creative at the end into a full-fledged creative direction tool.

Here are three tactics for putting this new generation of audio AI to work in your ads and content.

01

Build ads with emotion

A more natural voice? Pleasant. But it’s honestly the least interesting thing about Eleven v4.

The real shift is that you can now direct how each line lands, much like sitting in a recording booth with an actor. Minus the catering bill.

Audio Tags let you control emotion, pacing, delivery, reactions, and even some sound events right inside the script.

So instead of one flat voice-over, you build a mini-scene: curiosity → tension → panic → silence → relief.

Do it now:

  • Break your ad into emotional beats: Think in shifts of feeling, not just sentences.
  • Tag your key lines: Add tags before them, like [curious], [anxious], [whispers], [crying], or write your own more specific directions.
  • Stack directions when needed: Combine them, like [whispering, fearful], when one emotion isn’t enough.
  • Control pacing with punctuation: Use ellipses for longer pauses, dashes to cut thoughts short, and exclamation marks to add energy.
  • Generate several takes: Produce 3–4 interpretations of the same script and pick the one that fits your edit best.

Here’s our prompt. Tweak it for your own use case:

[peaceful, warm and conversational] Okay... I’m literally just here for one thing. [slightly amused] LUMIÉRA Glow Serum. [softly, pleased] I’ve been waiting ALL week to try this. [short pause] [read more…]

Listen to the audio.

Pro tip: Don’t tag every sentence. If the scene already carries enough context, v4 will hold the emotion and transition more naturally when you introduce the next change.

02

Suno v6: create a soundtrack around your finished video

We all know the workflow. You find a track that “kind of works,” then shuffle cuts around until the ad somehow bends to fit the music.

With v6, we flip that workflow on its head. Start with the finished visual and feed the video itself to the model as creative input for the soundtrack.

That gets especially useful in short ads, where the first three seconds, the product reveal, and the final packshot each need completely different energy.

Music stops being background filler and starts working with the edit.

Step by step:

  • Edit without music first: Cut the spot with only the dialogue and natural SFX in place.
  • Upload the video to v6: Use it as your reference material.
  • Describe the timing structure: Calm opening, gradual build-up, strong hit on the reveal, then a clean slowdown for the logo.
  • Ask for variations in energy: Request an instrumental version plus 2–3 alternatives with different energy levels, not just different genres.
  • Fine-tune in your editor: AI can follow the overall drama of the footage, but frame-perfect timing is still worth controlling by hand.

Here’s the prompt we used directly in Suno:

Create a 25-second soundtrack for the attached luxury watch commercial. The video already contains watch ticking, mechanical sounds, fabric, wind and city ambience. Keep them untouched. Create MUSIC ONLY around them. Style: dark luxury electronic, minimal cinematic, elegant, cold, precise, mysterious, expensive, restrained. Avoid pop, vocals, EDM, corporate music and heroic trailer sound. [read more…]

Pro tip: If you’re making music for ads, mind the licensing. Generate the final track you plan to use commercially on a plan that grants the appropriate commercial rights.

A great soundtrack loses its charm fast when legal gets involved.

03

Suno v6: one melody, multiple instruments

Want to see what Suno can really do? Don’t cram five melodies, seven genres, and a full orchestra into 30 seconds. That’s a demo of everything and nothing.

Do the opposite: keep one melody and change only the instrument.

The same few notes feel intimate on piano, dramatic on cello, monumental on brass, and aggressive on electric guitar. Same tune, entirely different mood.

The melody holds steady while the emotion shifts completely. Then, when every instrument comes back together for the finale, you get a natural, satisfying climax without forcing it.

Playbook:

  • Start with a short melody: Pick one that’s recognizable within 3–5 notes.
  • Map the arrangement in your prompt: Break it down by specific timestamps.
  • Lock the melody explicitly: State that it must stay the same, and that only the instrument, dynamics, and arrangement should change.
  • Use a reference for precision: If the motif needs to stay truly identical, provide it as an audio or MIDI reference instead of relying only on a text description.
  • Polish in Studio: Smooth out the transitions and, if needed, generate specific instruments as separate layers.
  • Bring it all home: Finish by reuniting every instrument on the same motif. It gives the listener a far more satisfying payoff.

Here’s how we structured our own prompt:

Create a 50-second instrumental piece built around ONE simple, memorable melody. The melody must stay clearly recognizable throughout the entire track. Do not invent a new theme for each section. Only change the instrument, texture and emotional character. [read more…]

Listen to the audio.

Sponsored by The Deep View

Become an AI expert in just 5 minutes a day

If you're a decision-maker at your company, you need to stay ahead of AI without spending your morning in webinars, reports, and half-finished podcasts. There is a faster way:

Subscribing to The Deep View.

This daily newsletter condenses the latest and greatest AI developments into a 5-minute read. They read the noise, pull out the signal, and explain what actually matters for work, strategy, and the tools your team is already using.

Trusted by 700,000+ readers at Google, Meta, Microsoft, and beyond.

Subscribe right here.

WEEKLY AI ROUNDUP

GOOGLE: Hallucinations got a mention in the updated gen-AI guidance, which now calls manual fact-checking of every AI-generated piece critical before you hit publish. And it’s not just body copy: title tags, meta descriptions, structured data and alt text all need the same human review.

META: Five million downloads in 22 days puts Muse among 23 apps ever to hit that uptake. And 3 million weekly users show people return. It stays free, with Meta taking a small fee on in-app transactions, so plan for shoppers who buy without visiting your site.

META: Advertising Week brought a pile of AI tooling: static images turn into video ads in Advantage+ Creative, Ads Creative Studio opens wider, and the business assistant goes agentic, generating creative and shifting budgets from a prompt. Plus in-chat checkout on Messenger.

OPENAI: Visual ads are coming to ChatGPT, appearing beside image generation for Free and Go users. Testing starts this month in the US with a limited advertiser group, alongside new attribution integrations and brand suitability pilots with DoubleVerify and IAS.

OPENAI: Invisible text watermarks are rolling out to ChatGPT and Codex output in the EU only, with opt-in watermarking for API customers worldwide. Detection is shaky, though: swapping just 10% of words for synonyms drops accuracy from 92% to 66%. Edit your drafts accordingly.

AI DISCLOSURE: Pennsylvania lawmakers have passed a bipartisan bill that requires advertisers to disclose AI use in consumer ads, joining a small group of states with similar rules. Start tracking which of your creatives use AI now, so you can label ads that run in Pennsylvania

Get the next AI tactic in your inbox.

Ready-to-apply AI tactics delivered every Saturday to help you get wins on Monday. Consumed in 7 minutes or less.

Keep Reading