Conversational Behavior and Rules

Defining the conversational behavior and operational rules of your AI chatbot is a critical step in ensuring that user interactions align with your brand’s voice and professional standards. While the chatbot’s identity provides a baseline persona, the behavior settings act as the governance framework, dictating how the AI processes queries, the length of its responses, and the boundaries of its knowledge.

• System Logic: How the AI interprets user intent.
• Response Constraints: Governing the technical depth and word count.
• Interaction Flow: Utilizing pre-built questions to guide the user journey.

Table of Contents

AI Persona and System Instructions

The core of your chatbot’s behavior is dictated by the AI Persona/Instruction field. This section allows you to provide high-level system instructions that function as the “laws” the AI must follow. Unlike a simple description, these instructions should include specific directives on what the bot can and cannot say. For instance, you might instruct the bot to “always refer users to the documentation for technical steps” or “never provide financial advice.”

To maximize the effectiveness of these instructions, it is recommended to integrate them with your existing Managing User Roles and Access in Docy: Best Practices and Plugin Recommendations strategies. By defining how the AI should prioritize information from your indexed documents versus general training data, you ensure a more accurate and site-specific user experience.

Managing Response Verbosity

The Antimanual plugin provides three distinct settings for Response Length, allowing you to tailor the output to the complexity of your subject matter and the preferences of your audience:

• Concise: Best for quick support widgets or mobile users. The AI will provide direct, short answers with minimal fluff.
• Balanced: The default setting, providing a mix of direct answers and helpful context, suitable for most business applications.
• Detailed: Ideal for technical documentation or complex troubleshooting where step-by-step explanations and thorough context are required.

Guiding Conversations with Pre-built Questions

Conversational behavior isn’t just about how the AI speaks, but also how it encourages user participation. By configuring Pre-built Questions, you can present visitors with suggested topics as soon as they open the chat interface. This reduces the “blank page” friction and ensures that users are guided toward your most valuable content or high-conversion pages.

When setting up these questions, consider the common pain points identified in your setup. Questions like “How do I install the plugin?” or “What are your pricing plans?” help streamline the support process and improve user satisfaction scores.

Language Support and Global Communication

In today’s global market, the Language Support setting is vital. You can define the primary language the chatbot uses to communicate, ensuring that the AI maintains consistent grammar and cultural nuance. Even if your site is multi-lingual, setting a primary communication language for the bot ensures that the system instructions are interpreted correctly within that linguistic context.

Frequently Asked Questions

Can I change the chatbot’s behavior rules after it is live?
Yes. Any changes made to the AI Persona or Response Length settings take effect immediately upon saving, allowing for real-time optimization based on user feedback.

What happens if a user asks something outside of my rules?
If you have provided strict “System Instructions,” the AI will politely decline to answer or redirect the user back to the permitted topics defined in your settings.

Does the response length affect my API costs?
Generally, yes. “Detailed” responses use more tokens from your AI provider (OpenAI or Google Gemini) than “Concise” responses. It is important to find a balance between depth and cost-efficiency.