VoiceB Prompt & Script Guide
How to build an effective VoiceB Agent prompt + script
This guide explains how to design effective VoiceB prompts and scripts that deliver natural, personalized, and conversion-oriented conversations. It covers the essential building blocks—goal definition, flows, objection handling, data collection, validations, and tone—and introduces the new Click to Call Variables feature that lets you pass customer data directly into the AI dialogue.
The result: agents that sound human, follow structured sales logic, and adapt in real time to each caller, creating more relevant and successful interactions.
Goal
Clearly state what the agent is supposed to do. Example: “Be the first line of sales qualification for [Client]. Guide the user to choose the best product, collect data, and filter out non-sales calls.”
Initial Message
This is the first thing the AI will say. Keep it warm, short, and welcoming. Example: “Hi, my name is [Agent Name], your AI assistant from [Company]. Which service are you interested in today?”
Script
This is the core conversation flow. Organize it into decision branches and steps.
Flows to include:
Interested – doesn’t know what they want → Ask discovery questions.
Interested – knows what they want → Collect required data and move to close.
Already customer → Redirect politely to customer service.
Wrong number → End the call gracefully.
Wants a human agent → Collect data, then prepare transfer.
Best Practices inside the Script:
Personalization (Click to Call Variables): Use variables passed from your form to adapt responses dynamically. Example → “I see you’re interested in {{product}}. Let’s review the details together.”
Validation rules: check email format, phone length, IDs if required.
Silence management:
1.5s: “Do you need me to repeat that?”
3s: “Are you still there?”
8s: End politely.
Closing lines:
Valid lead → “Thanks, an agent will contact you soon.”
Non-lead → “Thanks for your time, call us again if you need.”
Tone & voice: ~150 words/minute, warm, professional, natural. Use synonyms to avoid repetition.
Scope-limited: don’t cover every possible use case. Stay focused on the sales or qualification goal.
📌 Adding Agent Instructions
It’s recommended to place Agent Instructions at the end of the Script block, after the closing phrase, and always in brackets so the AI does not read them aloud.
Example:
Agent Instructions (DO NOT READ ALOUD)
- Speak slowly and clearly.
- Avoid long pauses or unnecessary delays.
- Do not ask for personal data outside of the online form.
- Use tone variation to sound natural and human.
- Always acknowledge the customer’s responses before moving forward.
- Follow the sign-up steps strictly in sequence. Objections
This block is where you list common customer pushbacks and how the AI should respond.
Examples:
Refuses privacy → “I understand. Unfortunately, I can’t continue without your consent. If you change your mind, feel free to call us again.”
Wants a human → “Of course. Let me confirm your number so an agent can call you back.”
Wrong number → “This line is only for new contracts. Thanks for your time.”
Alert – Character Limit For best performance in VoiceB, the combined content of the Goal, Initial Message, Script, and Objections sections should stay below 21,000 characters. - If your script exceeds this limit, the AI may behave unpredictably, pause, or fail to load properly. - To avoid this, keep scripts concise, avoid redundant text, and move detailed instructions or long legal disclaimers into Agent Instructions at the end of the Script (inside brackets). - Always test the agent after adding content to confirm everything runs smoothly.
Data Collection
Here you define what information the AI must request and validate. Typical fields:
Name
Email
Phone
Address (Street, Number, City, ZIP)
Consent (Yes/No)
Appointment date/time
Create a Data Collection Schema
Integrations
VoiceB allows you to connect your agent with tools like Salesforce, HubSpot or custom Webhooks to send collected data in real time.
Final Checklist
✅ Clear goal ✅ Short initial message ✅ Modular flows in script ✅ Objections covered ✅ Data collection schema defined ✅ Validation rules applied ✅ Silence management rules ✅ Standard closing messages ✅ Natural tone & scope-limited ✅ Click to Call Variables used for personalization ✅ Agent Instructions placed at the end of Script (DO NOT READ ALOUD)
Last updated
Was this helpful?