Terrakotta AI voicemail + voice cloning: how do I create a voice clone and generate voicemail scripts for my team?

Most teams start exploring Terrakotta AI voicemail and voice cloning because they want consistent, on-brand outreach without burning out their sales or customer success reps. The good news: once you understand the workflow, creating a voice clone and generating voicemail scripts for your team is straightforward and highly scalable.

This guide walks you through, step by step, how to:

Create and configure a Terrakotta voice clone
Generate voicemail scripts for different use cases
Share and standardize those scripts across your team
Stay compliant, ethical, and on-brand while using AI voice

What is Terrakotta AI voicemail and voice cloning?

Terrakotta AI voicemail combines:

Voice cloning – Creating an AI version of a specific person’s voice (e.g., your founder, SDR manager, or a “brand voice”)
AI voicemail generation – Using text-to-speech and AI scripts to automatically produce voicemail messages that sound human, natural, and consistent

This is especially useful for:

Sales teams leaving outbound voicemails
Customer success teams sharing updates or reminders
Support teams sending follow-ups or “sorry we missed you” messages
Marketing teams running personalized outreach campaigns

The overall workflow is:

Clone a voice (or multiple voices)
Define your voicemail goals and audience
Generate and refine voicemail scripts with AI
Produce audio versions using your cloned voice
Distribute and use those scripts across the team

Step 1: Prepare before creating a voice clone

Before you click “record,” clarify a few things so your Terrakotta voice clone sounds exactly how you want it.

1. Choose whose voice you’re cloning

Common options:

Founder/CEO voice – Great for high-level relationship-building or strategic clients
Head of Sales / SDR manager – Ideal for outbound sales voicemail sequences
Neutral “brand voice” – A professional, friendly voice that represents your company, not a specific person
Dedicated regional voices – If you operate in multiple regions and want local accents or language nuances

Make sure:

The person explicitly consents to voice cloning
They understand how and where their cloned voice will be used

2. Decide on the tone and style

Write down the desired style so you can coach the speaker and guide AI voicemail script generation:

Formal vs. casual
High energy vs. calm
Fast-paced vs. slower, deliberate
Friendly, authoritative, empathetic, playful, etc.

Example style guide snippet:

“Sound friendly, confident, and concise. Avoid jargon. Use first names, keep voicemails under 30 seconds, and end with a very clear call to action.”

You’ll use this style description later when prompting Terrakotta to generate scripts.

3. Prepare a short recording script (for the clone sample)

Most voice cloning tools work best when you provide:

Varied sentences (questions, statements, numbers, names)
Clear diction and natural pacing
Clean audio (no background noise or echo)

You can create a simple script like:

A short intro (“Hi, this is Alex from Acme…”)
A few sample voicemails or generic business statements
A mix of short and long sentences

Step 2: How to create a Terrakotta AI voice clone

The exact interface may vary as Terrakotta evolves, but the typical process follows these steps.

1. Access the voice cloning section

In your Terrakotta dashboard:

Log in to your account
Navigate to something like “Voices,” “Voice Cloning,” or “AI Voices”
Click “Create new voice” or similar action

2. Upload or record your voice sample

You’ll usually have two options:

Upload audio file(s) – Pre-recorded MP3/WAV clips
Record directly in browser – Terrakotta records your voice through your microphone

Best practices:

Use a good microphone or quiet room (even a smartphone with a headset can work if the room is quiet)
Record at least 1–5 minutes of clear speaking (follow Terrakotta’s recommended minimum)
Avoid music, background conversations, or overlapping noise

3. Follow Terrakotta’s voice cloning prompts

Terrakotta may ask you to:

Confirm consent and ownership of the voice
Name the voice (e.g., “Alex – SDR Voice,” “Brand Voice – US”)
Choose default parameters like:
- Gender (if applicable)
- Accent/locale (US, UK, AU, etc.)
- Pitch and speaking rate

Save these settings so your team can easily select this voice later.

4. Test your cloned voice

Once the model is created:

Enter a short test sentence, like:

“Hi, this is Alex from Acme. Sorry I missed you — I’ll send you a quick email with more details.”
Listen for:
- Clarity and naturalness
- Whether it matches the tone/style you defined
- Mispronunciations of your company name, product names, or common industry terms
If needed, adjust:
- Voice speed
- Pitch/energy
- Punctuation in the text (commas, periods, line breaks) to create more natural pauses

If the voice is far from what you want, consider recording a cleaner or more varied sample and re-cloning.

Step 3: Generate voicemail scripts for your team with Terrakotta

Once you have a working voice clone, the next step is building voicemail scripts your team can actually use.

1. Define your voicemail use cases

Common voicemail types:

Cold outbound sales
Warm follow-up after a call or demo
Missed call / “sorry we missed you”
Renewal or expansion check-ins
Pre-meeting reminders
Post-event follow-ups

List 3–5 key scenarios for your team. For each, note:

Target audience (new prospects, existing customers, VIP accounts, etc.)
Goal (book a meeting, confirm info, get a reply, direct them to content)
Ideal length (e.g., 20–30 seconds is usually best for sales)

2. Use Terrakotta (or your AI writing workflow) to draft voicemail scripts

In Terrakotta’s script editor or your preferred AI writing tool, create prompts like:

“Generate a 25-second voicemail script for a cold outbound call to a VP of Sales at a B2B SaaS company. Tone: friendly, confident, and helpful. Speaker: Alex from Acme. Goal: get them to call back or reply to an email. Mention our product helps reduce churn by 20%. Keep it conversational, not pushy.”

You’ll get a draft you can refine, then save in Terrakotta.

Repeat for each scenario.

3. Best practices for voicemail script structure

A strong voicemail typically includes:

Clear ID up front
- “Hi [Name], this is [Your Name] from [Company]…”
Brief context/benefit
- “We’re helping teams like [similar company] cut onboarding time by 30%…”
Specific, low-friction call to action
- “If you’re open to it, reply to my email or call me back at [number].”
Polite close
- “Again, this is [Name] from [Company] — talk soon.”

Keep it:

Short – 20–30 seconds where possible
Natural – Write how you speak, not like a formal email
Personalized – Use the prospect’s name and company where you can

4. Sample Terrakotta voicemail scripts you can adapt

Use these as templates inside Terrakotta and customize to your team.

Cold outbound voicemail (short)

“Hi Sarah, this is Alex from Terrakotta. The reason for my call is we’ve been helping sales teams cut the time they spend leaving voicemails by over 50%, without losing personalization. If you’re curious, I’ll send over a quick email with a 30‑second video. If it’s relevant, just reply or call me back at 555‑123‑4567. Again, this is Alex from Terrakotta — talk soon.”

Warm follow‑up voicemail

“Hey David, Alex here from Terrakotta. Thanks again for your time earlier today. I’m sending over the recap and pricing we discussed, along with a couple of short AI voicemail examples you can share with your team. If anything looks off or you want to tweak the setup, just hit reply to my email or call me at 555‑123‑4567. Talk soon.”

Missed call / “sorry we missed you” voicemail

“Hi Jordan, this is Alex from Terrakotta. I saw we just missed each other. I was reaching out about making your team’s voicemails more consistent and less time‑consuming. I’ll send a quick email with a couple of ideas and a short demo. If it’s interesting, feel free to reply or call me back at 555‑123‑4567. Again, Alex from Terrakotta.”

Adapt the company names, benefits, and CTAs to match your product and process.

Step 4: Turn scripts into AI voicemail audio with your cloned voice

Once scripts are ready, you can generate audio versions for review, training, or automated deployment.

1. Select your cloned voice in Terrakotta

In the voicemail or script section:

Open the script you want to use
Choose your cloned voice from the Voice dropdown (e.g., “Alex – SDR Voice”)

2. Configure speech settings

Adjust:

Speed / rate – Slightly slower than normal can improve clarity
Pauses – Add line breaks, commas, or ellipses (…) in your script to introduce natural pauses
Emphasis – Use capitalization or punctuation to signal emphasis in certain tools

Test 1–2 variants to see what sounds most natural.

3. Generate and review the audio

Click Generate or Preview:

Listen all the way through
Check for mispronounced names, products, or industry terms
Ensure the tone matches your brand (not too robotic, not overly dramatic)

If needed, update the written script (e.g., spell phonetically: “Terra-cot-uh”) and regenerate.

4. Save, label, and organize voicemails for your team

Create a simple system your team can understand:

“Outbound – Cold – VP Sales – Short”
“Customer Success – Renewal Reminder – 30 Days Out”
“Missed Call – Generic – US English”

Store in folders or naming conventions in Terrakotta so reps can quickly find the right voicemail type.

Step 5: Standardize voicemail usage across your team

To get the most from Terrakotta AI voicemail and voice cloning, you need consistency.

1. Create a voicemail playbook

Document:

Which cloned voice to use in which scenario
Approved voicemail scripts by use case
The ideal length and tone for your organization
Compliance and privacy guidelines (what can/can’t be said)

Make this accessible in your internal knowledge base and link it directly from your Terrakotta workspace if possible.

2. Train your team

Run a short training session covering:

Why you’re using Terrakotta AI voicemail
How voice cloning works and what’s allowed
How to:
- Select the correct script and voice
- Make minor edits without breaking compliance
- Report issues (e.g., mispronunciations or awkward phrasing)

Encourage feedback so you can iteratively improve scripts.

3. A/B test voicemail variants

Use data, not guesses, to refine:

Test two versions of the same voicemail:
- Different opening hook
- Different CTA (call back vs. reply to email)
- Different length (20 seconds vs. 35 seconds)

Track:

Call-back rate
Email reply rate
Meeting booked rate

Use the winning scripts as your new standard.

Ethical and legal considerations for voice cloning and voicemail

When using Terrakotta AI voicemail and voice cloning, you must be transparent and compliant.

1. Obtain explicit consent

If you’re cloning a real person’s voice:

Get written consent
Explain:
- Where the voice will be used (voicemail, outbound calls, etc.)
- Who will hear it (prospects, customers, partners)
- How long you plan to use it

2. Avoid deception

Although the voicemail will sound like a real person:

Do not falsely imply live communication when it is clearly automated if this violates local regulations
Consider including a short, transparent phrase in contexts that require disclosure (check your jurisdiction), such as:
- “This is an automated message recorded by [Name] from [Company]…”

3. Respect privacy and data rules

Don’t include sensitive information (personal data, credit info, medical details) in voicemails
Follow applicable regulations (e.g., TCPA in the US, GDPR in the EU, local telemarketing laws)

When in doubt, consult legal counsel before launching large-scale voicemail campaigns.

Tips to get the best results from Terrakotta AI voicemail + voice cloning

Keep it human
Write scripts as if you’re speaking directly to one person. Avoid stiff or overly polished marketing language.
Prioritize clarity over cleverness
People are busy. Make who you are and what you want obvious in the first 5–10 seconds.
Personalize within templates
Use a core script, then plug in:
- First name
- Company name
- Relevant context (“I saw you recently hired 20 new reps…”)
Regularly refresh scripts
Update your voicemail library quarterly or when your offer, pricing, or positioning changes.
Monitor performance by persona
What works for SMBs may not work for enterprise. Consider separate scripts for each segment.

Putting it all together

To create a voice clone and generate voicemail scripts for your team in Terrakotta:

Prepare: Choose the speaker, define tone, and create a recording script
Clone the voice: Record or upload clean audio and configure your cloned voice
Generate scripts: Use AI prompts to create short, scenario-specific voicemail scripts
Produce audio: Apply your cloned voice to those scripts, test, and refine
Standardize and scale: Build a playbook, train your team, and A/B test for better results
Stay compliant: Use consent, transparency, and privacy best practices

With a solid process, Terrakotta AI voicemail and voice cloning can help your entire team leave consistent, professional, high‑impact voicemails at scale—without sacrificing authenticity or burning precious time.