Calculate the time it takes for first response to a conversation in Intercom

Analyze the average time it takes your team to respond to a customer initiated conversation.

What do I get out of it?

Understand the performance of your customer support team and improve your customer satisfaction. You can have a breakdown by time cluster:
- % of conversations responded to within 24h
- % of conversations responded to within 48h


How does it work?

1) Import in a sheet the "Conversations" from Intercom. Call that sheet "Intercom conversations"

2) We'll use the columns "Statistics fist admin reply at" and "created at" to calculate the number of days between the time when the conversation was created and the time when somebody from your customer support team responded.

Let's add a calculated column called "Delta from creation to first reply" with the following formula: `=#Statistics fist admin reply at - #Created at`

3) Now you can write in any cell the following formula `=AVERAGE(#Delta from creation to first reply` to have the average response time.

Let's go further

4) Let's add a new calculated column called "Response time bracket" with the formula `=IFS(#Delta from creation to first reply<1, "<1 day", #Delta from creation to first reply<=3, "1-3 days", #Delta from creation to first reply>3, "+3 days")

5) Now let's build a dashboard on a new sheet called "Dashboard"

Write in B2, B3, and B4 the following values "<1 day", "1-3 days", "+3 days"

Then on C2, write the formula `=COUNTIF(Dashboard!#Response time bracket, B2)` and copy and paste the formula in C3 and C4

You now have the distribution of conversations per Response time time bracket. This is usually much more interesting than the average.

Indeed, an average of 1 day could mean you answer all your requests in exactly 1 day (which would be okay if you tell your customers you'll respond to them within a day) or that you respond to 50% of requests immediately and 50% after 2 days (which wouldn NOT be okay since you would underdeliver for half of your customers).

Of course, you could easily add more dimensions to this analysis: Month, week or customer support agent.