Tradesly · Guide
The reply thread: keep every response in context
Keep the customer’s replies together and step in with the context already in front of you.
A useful first text should open a conversation, not leave your team with a reply in a separate inbox. In Tradesly, open the lead’s Conversation panel to read the exchange and continue from the same business number when a person should take over.
The thread is a customer-facing history, so read it before writing. Confirm that you have the right person and the right job, especially when the customer has contacted the business more than once.
1 What your team sees
The Conversation panel shows the messages available for that lead or ticket. Use it to answer practical questions before stepping in:
- What did the customer ask for?
- What has already been promised?
- Is there a booking or next action to confirm?
- Did the customer ask for a person or express concern?
- Is an opt-out or other restriction visible?
Do not rely on an assumed match or on memory from another screen. If the identity or job context looks uncertain, verify it with the customer without repeating private details.
2 The steps
Open the conversation from your Office.
Search “Search by ticket ID, name, or phone…” or use the available sort options to find the person who needs attention.
Read the Conversation panel from the beginning of the current exchange.
Check what the customer requested, what was answered, and whether a next step is already shown.
Verify the recipient before sharing details.
If a phone number may be old, reassigned, or associated with more than one open request, keep the message generic until the person confirms who they are.
Write in the same thread when you need to step in.
Keep the response short, identify the business, and address the customer’s latest question. Honor every visible opt-out or contact restriction; do not try another channel to bypass one.
Add an internal note.
Use CSR Notes and Add note for context another teammate needs. Do not put internal diagnoses or unnecessary personal information in a customer-visible message.
Coach the customer-facing correction.
If the AI’s response should change next time, describe the desired outcome in its coaching workspace and test the revised behavior with your own number.
3 When a person should answer
Step in when the customer asks for a person, disputes a charge or promise, reports a safety issue, shares sensitive information, or needs an exception that is not clearly approved. The goal is a clean handoff with context, not to keep the AI in every exchange.
If a conversation is missing, attached to the wrong lead, or shows conflicting job details, stop messaging and contact support with the ticket or conversation identifier. Avoid copying the full transcript or customer phone number into ordinary email when an identifier is enough.
4 Keep the boundaries clear
Texting availability and timing depend on recipient permission, carrier status, your configuration, and applicable rules. AI-generated messages observe quiet hours. A human teammate can manually send outside those hours, so a staff-written message must be reviewed for both permission and timing before it is sent.