Every step in your sequence has a message attached to it. GetReplies gives you two ways to create messages: write them manually, or chat with the agent to have them generated for you.
Writing messages manually
Click on See Message for any step in the sequence to open the message editor.
The editor supports:
Plain text (recommended for all first emails and LinkedIn messages)
Basic formatting for email steps — avoid heavy formatting in cold outreach
Variables (inserted with a / slash command)
Signature (inserted with /signature — pulls the sender account’s saved signature)
Built-in variables
Type / in the message editor to open the variable picker.
Built-in variables are automatically populated from your contact list:
Variable | How to insert | Example output |
First name | Type /first name | "Hi Sarah" — contact’s first name |
Last name | Type /last name | "Smith" — contact’s last name |
Companny name | Type /company name | "Acme Corp" — company from contact record |
Job title | Type /job title | "Head of Marketing" — job title |
LinkedIn headline | Type /linkedin headline | Full LinkedIn headline of the contact |
Signature | Type /signature | Sender account’s saved email signature |
Any custom field | Type /field_name | Any custom column uploaded in your CSV |
Custom variables from your CSV
Any column you mapped as a custom field during CSV upload appears in the variable picker. This is how you use research data in your messages.
For example: if your CSV has a column called pain_point mapped as a custom field, you can insert this variabl into your message. Every contact receives a message with their specific pain point filled in automatically.
Chatting with the agent to generate messages
Instead of writing messages manually, you can chat with the agent in the sequence panel. The agent asks questions about your product, ICP, and CTA, then generates all the messages in your sequence automatically.
This works best when your Knowledge Base is populated. The agent draws from your product entry, case studies, and event entries to write messages that are relevant and specific.
You can always edit agent-generated messages manually. Think of the agent output as a first draft — usually 80–90% of the way there, needing only minor refinements in tone or specificity.