Close the loop with customers by turning survey feedback into meaningful responses. With Survey Response Agent, you can automate replies to survey submissions at scale and set up different response workflows based on customer feedback.

Choose how you want each response handled - let AI reply autonomously, or route AI-generated replies for human approval before they're sent to the survey respondent.

From the left navigation, go to Survey AI.

Click the Agents drop-down arrow, then select Survey Response Agent.

Click Library.
​

It displays prebuilt agent templates for different survey response workflows.
​Choose an agent template:

Handy Tip: Choose the template that best fits your workflow. You can view the complete agent flow before using it.

Click Use Agent, then rename the agent as needed.

Create a survey that you want the agent to respond to.

In the agent's Trigger, select the survey you created.

Configure the agent workflow. Review the agent details and configure the trigger, tasks, conditions, response generation, and how the generated response will be handled.

Publish and test the agent. Save your workflow, click Publish, and submit a response to the survey you selected in the trigger to test the workflow.
​

After you click Use Agent, the agent workflow opens. From here, you can review the agent details, test the workflow, and configure its trigger, tasks, and controls.
​

Click the pencil icon to open the details drawer on the right.

Here, you can review and update the agent's Name, Goals, Outcomes, and Locations. These fields are populated by default, but you can update them to align with your requirements.
​

Click Save after making your changes.

Use the Preview icon at the top of the screen to review the agent settings and test the workflow before publishing it.

After testing, you can:

Use the icons on the left to work with different parts of the agent:

Use the Fit to screen option to view the complete workflow on your screen.

The workflow starts with the New Survey Response trigger.

Select the survey you want the agent to respond to from the drop-down list. The agent runs whenever a customer submits a response to the selected survey.

Add conditions to determine how different survey responses should be handled. For example, you can create separate branches for:

You can create multiple branches and conditions within the same workflow to handle different types of customer feedback.

When you're done, click Save and add a description for the workflow.

The agent can use a customer's previous survey responses when generating a reply. This gives the agent additional context to understand the customer's feedback and create a more relevant response.
​

You can review or modify the Task name, Description, and Tools as needed. When you're done, click Save.

For example, if a customer previously submitted positive feedback but later submits a negative response, the agent can use their response history to understand the change in their experience and generate a more relevant reply.
​

This helps the agent respond based on the customer's feedback history instead of treating every survey response as an isolated interaction.
​

The Survey Detail Extraction task analyzes the information provided in the survey response and extracts key details that help the agent understand the customer's feedback and generate a more relevant reply.
​

Because surveys can include different question types, the agent can identify and analyze details such as:

The extracted information gives the agent a clearer understanding of the respondent's feedback and provides additional context for generating a personalized and relevant reply.
​

Handy Tip: You can modify the Task name, Description, LLM model, Input field, System prompt, User prompts, and Output field as needed.
​

The agent uses the extracted survey details and available customer context to generate a personalized reply.
​

It considers the respondent's feedback, previous survey responses, contact history, and brand voice to create a relevant response.
​

You can also customize the instructions used to generate the response to add specific requirements, guidelines, or nuances based on your workflow.
​

The final survey response is assembled using the drafted response strategy, extracted survey details, contact history, and brand voice.
​

Handy Tip: You can modify the Task name, Description, LLM model, Context, Input fields, System prompt, User prompts, and Output fields to customize how the response is generated.
​

Once the agent generates a reply, choose how you want it to be sent to the survey respondent:

You can also use conditions to apply different actions to different types of responses. For example, you can send Promoter responses automatically while routing Detractor responses through a human approval workflow.
​

This gives you the flexibility to automate responses where appropriate while keeping a Birdeye user in the loop when a response requires additional review.

Once you've configured the workflow, click Save to save your changes.

Click the highlighted icon to test the workflow.

Now, click Run test.

Click Publish to activate the agent.

Submit a response to the survey you selected in the Trigger to test the complete workflow and verify that the agent handles the response as expected.

If you selected Send after approval, the generated response appears in the Approval tab under Human Action.
​

A Birdeye user can review the response and:

This gives your team control over responses that require human review while still using AI to analyze survey feedback and draft the reply.
​

With Survey Response Agent, you can automate responses to customer feedback at scale while giving your team control over when AI can respond independently and when human review is needed.