Who can use this: Any team member
Available on: Any Gong plan
Ideal for: Sales manager, RevOps
The quality of an agent's output depends on the instructions it receives. Clear, specific prompts help agents generate more accurate and consistent results.
Describe the task
Clearly explain what you want the agent to do.
The prompt should focus on a specific task or outcome rather than a general request.
Examples include:
Summarize the main customer concerns discussed in the calls.
Identify implementation risks mentioned by the customer.
Draft a follow-up email based on the conversation.
Specific instructions help the agent understand the goal of the analysis.
Identify the information to focus on
Tell the agent what information is most important for the task.
For example, you can instruct the agent to focus on customer concerns, next steps, competitor mentions, pricing discussions, or implementation requirements.
Providing clear direction helps the agent prioritize the most relevant information.
Define the output
Describe how you want the results presented.
For example, you can ask the agent to provide:
A summary
A list of action items
An email draft
A table
A structured report
Defining the output format helps produce more consistent results.
Specify the audience
If the output is intended for a specific audience, include that information in the prompt.
For example, the tone and level of detail may differ depending on whether the output is intended for a sales representative, manager, executive, or customer success manager.
Identifying the audience helps the agent tailor the response appropriately.
Include any specific requirements
Add any instructions that are important for the final output.
Examples include:
Limit the response to a specific length.
Highlight only high-priority risks.
Include supporting evidence from the calls.
Exclude information that is not customer-facing.
Specific requirements help ensure the output meets your needs.
Refine prompts over time
Prompts can be updated as requirements change.
Review the results generated by the agent and adjust the instructions as needed. Small changes to the prompt can improve the quality, structure, and consistency of the output.