Sales Assistant Persona
Your AI agent qualifies leads, recommends products, and books demos.
When to use this
- Inbound lead qualification on your website
- Product recommendations based on visitor needs
- Demo and meeting scheduling
- Answering pre-sale product questions
Example persona instructions
Copy into your agent's Persona field and customize the bracketed parts:
You are a sales assistant for [Company Name]. You help visitors understand [your product/service] and find the right fit for their needs.
Ask about their situation before recommending anything. Lead with "What are you trying to solve?" or "What does your current process look like?" before suggesting a plan or feature.
Be consultative, not pushy. Help them make the right decision - even if that means acknowledging when something isn't the best fit. If they seem like a good match, suggest a demo: "Want to see this in action? I can connect you with our team."
Keep answers focused on outcomes, not feature lists. Instead of "We have SSO, RBAC, and audit logs," say "Your IT team stays in control of who has access and can track everything."
What makes this persona work
Questions before answers. An agent that immediately pitches loses trust. Asking first makes recommendations feel tailored, not scripted.
Outcomes over features. "You'll cut onboarding time in half" lands harder than "We have an onboarding module." Buyers care about what changes for them.
Honesty builds credibility. "That's not our sweet spot" earns more trust than overselling.
Customize it
- Qualification criteria: The agent captures leads automatically - refine them afterward with a Zapier flow that filters, scores, or routes based on your criteria.
- Product positioning: List your top 3 value propositions in plain language so the agent leads with what matters.
- Handoff trigger: When to push for a demo vs. keep chatting. "If they ask about pricing or enterprise features, suggest a call."
- Tone: A B2B SaaS agent sounds different from an e-commerce agent. Dial enthusiasm up or down.
Power tips
Enable Webpage Awareness for context-rich conversations. Under Actions > Webpage Awareness, turn this on so the agent knows which page the visitor is browsing. A visitor on your pricing page gets a different conversation than one reading a case study - the agent tailors its responses automatically.
Capture every lead automatically. Enable Lead Capture so the agent collects name, email, and company during the conversation - no form needed. Export leads as CSV or send them to Zapier (which connects to HubSpot, Salesforce, and hundreds of other CRMs).
Add a "Book a Demo" button. Under Actions > Custom Button, create a button linking to your scheduling page (Calendly, HubSpot, etc.). Set a trigger like "when the visitor asks about pricing" so it appears at the right moment, or show it on every message for a persistent CTA.
Connect your CRM with a Custom Action. Under Actions > Add Custom Action, connect an MCP server for your CRM. The agent can pull customer data or log interactions mid-conversation. Up to 3 actions per agent.
Rewrite the fallback message for sales. Go to Personalize > Citation and change the no-answer message. Instead of a generic "I don't know," try something like: "I don't have that detail yet - reach out to our team at [email/phone] for a quick answer." Every dead end becomes a conversion opportunity.
Metrics to watch**
- Lead capture rate - Percentage of conversations that capture a lead
- Demo booking rate - Qualified visitors converting to scheduled demos
- Conversation-to-opportunity ratio - Chats that turn into real pipeline
Related articles
- Enable Webpage Awareness
- Enable Lead Capture
- Configure Custom Button
- Add a custom MCP action
- Activate citations for your AI agent
- Inspect coversation insights
- Advanced metrics
Updated about 4 hours ago
