### Answer You can configure your outbound agent to wait for the customer to speak first by setting the conversation initiation option to "User-First" in the agent configuration [(1) ]() . This setting makes the agent wait for user input instead of initiating the conversation . However, if you want to add a delay before the agent speaks (for example, to give the customer time to raise the phone to their ear) , you can use the

begin_message_delay_ms

parameter [(2) ]() [(3) ]() [(4) ]() . This delays the first message by the specified amount of milliseconds, with a valid range of [0, 5000] . Note that this only applies when the agent speaks first . One workaround mentioned in the community is to set the AI to speak first but use just a dash "-" as the initial message [(5) ]() . This makes the agent think it has spoken, but it remains silent, and then the conversation proceeds normally after the user speaks .

​ To learn more, see the sources I considered: [[1] how to make voice ai on outbound to talk less](https://discord.com/channels/1186424376300679209/1314325013876183091) [[2] Create Voice Agent](https://docs.retellai.com/api-references/create-agent) [[3] Update Voice Agent](https://docs.retellai.com/api-references/update-agent) [[4] Get Voice Agent](https://docs.retellai.com/api-references/get-agent) [[5] Discord message from @bajacheebo on 2024-12-14](https://discord.com/channels/1186424376300679209/1186462578310910112/1317317835734908968) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer You're correct—Conversation Flow agents do have the

start_speaker

parameter to control who initiates the conversation [(1) ]() [(2) ]() [(3) ]() . To make your outbound Conversation Flow agent wait for the customer to speak first, set

start_speaker

to

"user"

. This parameter determines who starts the conversation and accepts either

"user"

or

"agent"

as values . Here's how to configure it:

javascript
const conversationFlowResponse = await client.conversationFlow.create({
  start_speaker: 'user',
  // ... other configuration
}) ;

[(3) ]() If you want the agent to wait for a specific duration before starting (when configured as agent-first) , you can use the

begin_after_user_silence_ms

parameter [(1) ]() [(2) ]() [(3) ]() . This makes the AI begin the conversation after waiting for the user for the specified duration in milliseconds . However, this only applies if the agent is configured to wait for the user to speak first . **

​ To learn more, see the sources I considered: [[1] List Conversation Flows](https://docs.retellai.com/api-references/list-conversation-flows) [[2] Update Conversation Flow](https://docs.retellai.com/api-references/update-conversation-flow) [[3] Create Conversation Flow](https://docs.retellai.com/api-references/create-conversation-flow) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer You should add

start_speaker: 'user'

when you create or update your Conversation Flow in Retell [(1) ]() [(2) ]() . This is configured at the Conversation Flow level, not when you make the outbound call. Here's where to set it: When creating a new Conversation Flow:

javascript
const conversationFlowResponse = await client.conversationFlow.create({
  model_choice: { model: 'gpt-4.1', type: 'cascading' },
  start_speaker: 'user',
  nodes: [
    {
      id: 'start',
      instruction: { text: 'Greet the customer and ask how you can help them.', type: 'prompt' },
      type: 'conversation',
    },
  ],
}) ;

[(1) ]() When updating an existing Conversation Flow:

javascript
const conversationFlowResponse = await client.conversationFlow.update(
  'conversation_flow_id',
  {
    start_speaker: 'user',
    // other configuration
  }
) ;

[(2) ]() Once you've configured

start_speaker: 'user'

in your Conversation Flow, any agent using that flow will wait for the user to speak first—including outbound calls [(1) ]() [(3) ]() [(2) ]() .

​ To learn more, see the sources I considered: [[1] Create Conversation Flow](https://docs.retellai.com/api-references/create-conversation-flow) [[2] Update Conversation Flow](https://docs.retellai.com/api-references/update-conversation-flow) [[3] List Conversation Flows](https://docs.retellai.com/api-references/list-conversation-flows) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer I understand

start_speaker: 'user'

isn't working as expected. Let me help troubleshoot this. First, verify your configuration: You can configure

start_speaker

in the Dashboard or via API [(1) ]() . In the Dashboard: Click on the

begin

icon in your Conversation Flow, and you can select who speaks first in the call . Via API: When creating or updating your Conversation Flow, ensure

start_speaker

is set to

"user"

[(2) ]() [(3) ]() [(4) ]() :

javascript
const conversationFlowResponse = await client.conversationFlow.create({
  model_choice: { model: 'gpt-4.1', type: 'cascading' },
  start_speaker: 'user',
  nodes: [
    {
      id: 'start',
      instruction: { text: 'Greet the customer and ask how you can help them.', type: 'prompt' },
      type: 'conversation',
    },
  ],
}) ;

[(4) ]() Additional consideration: For outbound calls specifically, you might also want to configure the "Pause before speaking" setting [(1) ]() . This setting controls how long the agent waits before speaking at the beginning of the call if the agent speaks first, which is useful to handle scenarios when the user is still picking up the phone . Workaround from the community: If

start_speaker: 'user'

still doesn't work, one community member mentioned setting the

begin_message

to an empty string

""

to make the agent wait for the user to speak first [(5) ]() . This approach was documented for Retell LLM agents [(6) ]() [(5) ]() . Can you confirm: 1. Where you set

start_speaker: 'user'

(Dashboard or API) ? 2. Whether you've verified the Conversation Flow was successfully updated with this setting? 3. If you're seeing any specific error messages or unexpected behavior?

​ To learn more, see the sources I considered: [[1] Step 1: Configure global settings](https://docs.retellai.com/build/conversation-flow/global-setting) [[2] List Conversation Flows](https://docs.retellai.com/api-references/list-conversation-flows) [[3] Update Conversation Flow](https://docs.retellai.com/api-references/update-conversation-flow) [[4] Create Conversation Flow](https://docs.retellai.com/api-references/create-conversation-flow) [[5] Discord message from @gianni0305 on 2024-04-06](https://discord.com/channels/1186424376300679209/1186462578310910112/1226254960497524929) [[6] Update Retell LLM](https://docs.retellai.com/api-references/update-retell-llm) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.