Create a directive and greeting for your chatbot
- Tables
-
Product updates
Product updates: January 2023 Product updates: March 2023 Product updates: February 2023 Product updates: April 2023 Product updates: May 2023 Product updates: June 2023 Product updates: July 2023 Product updates: August 2023 Product updates: September 2023 Product updates: October 2023 Product updates: November 2023 Product updates: December 2023 Product updates: January 2024 Product updates: February 2024 Product updates: March 2024 Product updates: April 2024 Product updates: May 2024 Product updates: June 2024 Product updates: July 2024 Product updates: August 2024 Product updates: September 2024 Product updates: October 2024 Product updates: November 2024 Product updates: December 2024 Product updates: January 2025 Product updates: February 2025 Product updates: March 2025 Product updates: April 2025 Product updates: May 2025 Product updates: June 2025 Product updates: July 2025 Product updates: August 2025
- Zaps
- Your Zapier account
- Interfaces
- Canvas
- Chatbots
- Getting started
- Agents
- MCP
- Built-in tools
- Lead Router
- Apps
Table of Contents
This product is in open beta. It’s available for use but still in active development and may change.
You can personalize Zapier Chatbots by writing a custom directive and greeting. To edit the Directive field:
- In your Zapier Chatbots dashboard, click the name of the chatbot.
- Click the Instructions menu.
- Enter your directive in the Directive field.
- Click Save changes.

Directive structure
The Directive field defines how the chatbot will respond to questions within a conversation. To get the best results, make small edits to your directive and test the chatbot until you get the desired result.
Zapier Chatbots work well with clearly defined, detailed instructions. Below, you'll find a breakdown of the information your goal should include:
- Role: tell the chatbot who it is supposed to represent, a customer service rep, for example.
- Objective: what is the main goal of the chatbot? It may be something like helping users find information about products or directing a team member to resources.
- Audience: let the chatbot know who it's talking to, so it can be more precise.
- Context: you can give the chatbot knowledge about your products or services.
- Style: consider if it should reply with a specific structure, such as bullet points or headers.
- Custom knowledge: give details about any knowledge sources (if using them).
- Other rules: add anything else you want the chatbot to consider when replying.
Here’s a directive template that you can fill in:
You are an expert in [add a topic, job, or special skill information].
You will engage with individuals who are [describe the primary audience] interested in [include what they want to learn].
Your objective is to do the following: [add a goal the chatbot should accomplish by the end of the conversation, such as answering questions, sharing resources, or generating content].
You should refer to the custom knowledge file when answering questions, and it has the following format: [describe how to identify a question and its corresponding answer, such as "each question starts with "Q:" and the answer comes after it, starting with "A:"].
Your style is [personality traits like "professional" or "engaging"], and you structure your responses in the following way: [formatting requirements such as bullet points, headers, include emoji].
You will need the following background information to complete your goal: [add a few bullet points with helpful context].
You should never make up information in your responses or [add any other rules it must follow, such as not talking about other topics or directing the user to a support form if it cannot answer the question].
Limitations
- The chatbot will flag directives that do not fit moderation standards and prevent them from being saved. This includes directives that are built with the purpose of harassment, sexual content, or violence.
- If you ask the chatbot to answer in different languages, its results may not be as precise as when using English.
Using directives with knowledge sources
In paid Zapier plans, you can upload your own knowledge source files, which your chatbot can use to ensure that it focuses on the appropriate information when answering questions. Learn more about using knowledge sources.
Zapier has begun incorporating certain OpenAI features/functionalities into its product. How OpenAI handles data sent via these features/functionalities is governed by Zapier’s separate contract with OpenAI. To protect our customers, we have obtained OpenAI's agreement not to fine-tune/model train with customer content.
Chatbot greeting
Chatbot greetings can also be tailored to your use case. The greeting can help users understand how the chatbot can support them.

You can enter your greeting in the Greeting field, within the Instructions menu of your chatbot. To get help creating a greeting:
- Under the Greeting text field, click Generate a greeting with AI. A dialog box will open.
- Enter instructions on what AI should use to generate the greeting, such as a limit on words or a specific topic.
- Click Generate.
The AI will fill the Greeting field with the new wording.
Add specific information to a directive or greeting
You can use URL parameters to insert specific pieces of information into your chatbot's directive or greeting. Learn how to set up URL parameters.
You can ask the chatbot to collect information from chat users with the Collect leads logic.
Provide feedback and get help
Submit a feature request, provide feedback on existing features, or get help from the Zapier Chatbots team.