Skip to main content

 

Hey All!

 

I have extracted the conversation data using Intercom REST API (link) to our data warehouse. The use case is that I want to build a First response time (Bot inbox time excluded) metric from the Intercom reporting in our tableau report. However, we are not sure which fields to use for the calculation. I do have a field named  time_to_admin_reply but it includes bot inbox time. Additionally, I am not extracting conversation parts as it is very time-consuming.

 

It would be great if anyone could help me with this. Also, let me know if more information is needed. Thanks in Advance.

Hi @bshir 😄  I’ll leave the side of exporting things open to someone who may have implemented something similar, but in terms of excluding bot response times from FRT reports, you can achieve this with the following steps:​
If you want to exclude bot responses from that Metric, you'll want to enable the Bot Inbox first. To enable the bot inbox, go to Automation > Settings and open Turn on the bot inbox to toggle this on:

Screenshot+2023-06-28+at+16.30.01.png?expires=1716558300&signature=e340283c22690151f20709ed31837a6bbf902bb8a4af2fbded6c2d23770af757


With the dedicated bot inbox enabled, the Time to close and First response time metrics in your Inbox reports will have a new filter you can apply to include or exclude the time bots spend handling a conversation.

With the bot inbox enabled, you’ll see a new filter on the following Inbox reports:

Note: It’s only possible to exclude bot inbox time from your reports for the period since you enabled the bot inbox.

You can find more information on that in the article below!

Turn on the bot inbox


Thanks for your response @bernice !

 

Everything is in order at Intercom report side, we are monitoring FRT excluding bot response time using Intercom reports for a while now.

The issue is that we want to calculate the same metric outside of Intercom (in our Tableau reports), for which we have extracted the conversation data using Intercom REST API (link). But we are not sure which fields to use for the calculation.


Hi @bernice - we are also interested in the follow up posed by @bshir - we’d like to calculate this metric outside of Intercom using the API if you could please advise. Thank you.


Hi @Dylan Seidt & @bshir! To calculate the First Response Time (FRT) using API data, you should measure the time difference between the first user message and the first admin reply in a conversation. You can use the created_at or first_contact_reply_at timestamps for the first user message and the timestamp of the first admin reply to calculate this difference.

Keep in mind that the FRT metric measures the first teammate response after the conversation is assigned to a teammate, but it still counts the response time from the start of the conversation. If you want to measure the Teammate First Response Time, you can find it in the Teammate Performance Report, but this is more difficult to measure via API as it applies per teammate, not per conversation. You would need to know the teammate ID and check which teammate's first response time for that conversation is accurate.

Hope this helps! Let me know if you have any further questions on this🚀


Hi @bernice - I am specifically looking for a way to exclude bot time. My understanding is created_at and first_contact_reply_at are the timestamps of the user’s first message. I’m looking for the timestamp of when the user requested human assistance. Here’s the filter I am using on the responsiveness report:

And here is the table I am hoping to reproduce:

 


Hi Dylan Seidt, It’s Mat from the Support Engineering Team 😀

When you go to Reports and try to create a new graph, you can select the FRT metric that excludes the bot reaction. Look at the screenshot below👇

 


To calculate First Response Time in Intercom excluding bot inbox time, filter interactions by "First response" and exclude bot-generated responses. Use reporting tools to determine the average time between the customer's initial message and the first human response. (https://www.hqlsolutions.com/)

 

 

 

 


Reply