BusinessChat 's SLAs reports show how fast the team responds to customers and resolves conversations. Managers can now spot delays, compare agent performance, and improve customer experience with real data. First Response Time, Average Response Time, and Team Resolution Time all measure how long customers wait, but each starts and stops its clock at a different moment.
Overview: Three Clocks, Three Questions
Every report measures a stretch of time and shows you the average. The difference is where the stopwatch is pressed:
Report | What it Answers | Clock Starts | Clock Stops
|
First Response Time | How long until someone picks this up and says something? | The conversation is first assigned to a person. | That person's first reply. |
Average Response Time | Once we're talking, how quickly do we keep replying? | The customer's first message after your last reply (or when a person is assigned). | The next reply from a person. |
Team Resolution Time | How long does the whole thing take to finish? | The bot hands the conversation over to a team. | The conversation is closed and rated "Resolved". |
Ground Rules: What All Three Reports Have in Common
Time Limits: All three refuse a date range longer than 31 days.
Timezones: Days, weeks, and months are cut on your business clock, not UTC.
Bot Exclusions: Choosing "Bot" in the Team-or-User picker empties all three, as none count bot or AI activity.
Comparisons: The percentage chip compares your range against the identical span immediately before it, showing green when waiting times fall.
Valid Replies: A waiting customer stops waiting when the assigned person sends a message, reassigns the conversation, or closes the conversation. Handing a conversation on ends the customer's wait on that person.
Working Hours: Every card has a switch to toggle between 24/7 (plain wall-clock time) and Working hours only (counting only minutes inside configured team shifts).
1. First Response Time
The First Response Time report measures the average time from assignment to the first reply. It captures the gap between a conversation landing on a person's plate and that person saying their first word.
Calculation: Every measured first reply in your date range is added up and divided by the number of replies, giving each reply equal weight.
Exclusions: It leaves out conversations still waiting for their first reply, bot and AI replies, and everything before a person was assigned.
Example: The bot hands a conversation to Maya at 10:00. Maya sends her first message to the customer at 10:06. The First Response Time is 6 minutes, credited to Maya and her team.
2. Average Response Time
The Average Response Time report measures the average time to reply each time a customer is left waiting across the entire conversation.
Multiple Messages: Several customer messages in a row count as a single wait, with the clock starting at the first of them.
Credit: Each reply is credited to the person who actually sent it, meaning colleagues stepping in mid-conversation show up correctly.
Exclusions: Waits that are still open, and bot or AI replies are excluded.
Example: A conversation is assigned to Maya at 10:00, and she replies at 10:06 (Wait 1 = 6 min). The customer writes at 10:20 and 10:21. Maya replies at 10:35 (Wait 2 = 15 min). The customer writes at 10:50, and Maya closes at 11:00 (Wait 3 = 10 min). The Average Response Time is 10 min 20 s based on 3 counted replies.
3. Team Resolution Time
💡 Note on Availability: While the First Response Time and Average Response Time reports are available on all accounts, the Team Resolution Time report requires an active Conversation Success Index (CSI) subscription. This is because the resolution report relies directly on CSI conversation ratings to function.
The Team Resolution Time report measures the average time from handover to your team until the conversation is resolved. Only Admins and Owners can see this report.
This report strictly measures the resolution time from the moment the conversation is handed to a human employee until it is closed.
Resolved Only: It only counts sessions the Conversation Success Index rated as "Resolved".
Settings: Managers can toggle to Include or Exclude customer wait times. Excluding it pauses the clock while you wait for a customer to reply, showing your own true handling time.
Credit: The whole duration goes to the person and team who owned the conversation when it was closed, even if it passed through several hands.
Example: Maya takes a chat at 09:00. Between 09:00 and 13:00, there are several exchanges, with 2 h 35 m spent waiting on the customer. Maya closes it at 13:00, rated Resolved. On 24/7 including wait, the time is 4 hours, but excluding wait, it is 1 h 25 m.
Exporting Data
Each of the three SLAs charts has an export control in its header to instantly download a CSV file. The file is built from the figures on your screen.
Response Reports Exports: First and Average Response Time exports contain four columns: Day/Week/Month, Replies, 24/7 Average time (seconds), and Working hours Average time (seconds).
Resolution Report Export: Team Resolution Time exports contain: Day/Week/Month, Resolved sessions, With wait Average time (seconds), and Without wait Average time (seconds).
Data Accuracy: The CSV keeps decimals for unrounded seconds so your own SLA calculations stay accurate.
Missing Periods: Empty periods are absent from the file entirely. A blank cell means that reading had nothing to measure.
To download one of the SLAs reports, go to Reports from the main BusinessChat dashboard > Click on Conversations, under Engagement reports > Loacte the SLA report you need and click the three dots and download.
Frequently Asked Questions (FAQs)
Why doesn't the headline match the average of the chart points? The headline weighs every reply equally across the whole date range, whereas each chart point is an average only for that specific period (like a single day).
Why is a day missing from the chart and export? Nothing was measured in that period. Writing a zero would incorrectly read as "answered instantly".
Why did numbers change when switching to "Working hours only"? Time outside shifts is removed, and any wait that fell entirely outside shifts drops out of the calculation completely.
A conversation took all day but response times look fine. Why? Response times measure the gaps between messages, not the total length. Quick replies strung out over a whole day yield great response times but poor resolution times, which is why Team Resolution Time is useful.
Why don't I see Team Resolution Time? Your role is not Admin or Owner, or Conversation Success features are not enabled.
I picked "Bot" and everything emptied. Is this broken? No. All three reports measure what people did; bot and AI messages never start or stop these clocks.