Insights recipe: Reviewing SLA performance in Insights

Return to top
Have more questions? Submit a request


  • Sarah Ryan

    Hi Veronica, To change the graph type to a table you'll just click the Table Icon, which is 2nd graph type icon.

  • Christopher Rehn

    Hi. Struggling a bit with these metrics.

    Having copied verbatim the 4 example custom metrics and then applying the filter Week (Mon-Sun)/Year (Ticket Solved), my total adds up to 4648 tickets. If however I use the WHAT # Tickets Solved, I get 6893 tickets. What may cause this discrepancy?

    The same applies if I select Tickets Created for filter rather than Ticket Solved, also for last updated.

    I can confirm that the 4 metrics I applied are worded exactly like the ones stated at the top of this article.

    Could the difference stem from tickets being solved without a public comment from an agent?

  • Amy Dee

    Hi Christopher! You got it! The difference is almost certainly coming from tickets that do not have a public reply.

    The metrics described above will only count tickets that have a value for First Reply Time. If a ticket was solved without any public agent comment, it will not be included in a first reply percentage, but it will be included in a general ticket count.

    I hope this helps! Happy reporting!

  • Christopher Rehn

    @Amy: Thanks for the swift reply.


    Is there a reliable way to create a report providing "Solved tickets without a first reply", a way to see the number of tickets being solved without providing the customer with a public comment?

  • Amy Dee

    Hi Christopher! There are a number of ways to approach that sort of report. For now, let's use the method that most closely matches the recipe above. 

    In Step 3, the recipe uses this metric to find tickets with at least one agent reply:

    • SELECT # Tickets WHERE Ticket replies > 0

    To see tickets without a reply, set replies equal to 0:

    • SELECT # Tickets WHERE Ticket replies = 0

    In Insights, the Ticket replies fact shows a count of public agent comments after ticket creation. If this fact is 0, then the ticket has no first reply. (The ticket creation comment does not count as a reply, even if it was submitted by an agent.)

    You can use this metric on its own; you do not need to build it into a percentage like the recipe does. It will behave just like the default # Tickets metric, but with filtered results.

    I hope this helps! Happy reporting!

  • Schuyler Weinberg

    Is there a way to report on tickets that are not replied to outside of a different SLA metric such as Next Reply Time?

  • Jessie Schutz
    Zendesk team member

    Hi Schuyler! Welcome to the Community.

    I'm not sure I totally understand your question...Can you elaborate on what you're trying to do here?

  • Andrew Sapienza

    I'm having a bummer of a time relabeling the metrics. 

    Here is what it looks like currently:

    Also, this is pretty unclear to me:

    > To change the label all you need to do is go to the table format of the report, double-click that header, type in a new label (I like changing it to a simple 1-8 hrs format), and hit enter. Repeat this step for each of your metrics if you wish.

    Nothing I click on gives me an editable field. Please help! 

    One other thing, what if I wanted to group the Biz and Cal hours by time to first reply? It's what I'd prefer to do.  

  • Lettie Ball


    I have created the above custom metrics for a FRT < 48 hours and greater than 48 hours. 

    I wanted the metrics in hours, so I did the below: 


    Select #Tickets WHERE First reply time in minutes/60 <=48  

    Select #Tickets WHERE First reply time in minutes/60 >=48  


    I'm a bit concerned about the report as it displays incredibly high values: 


    Have I done something wrong with my custom metric by trying to make it hours as opposed to minutes? Also is there a way to filter this report to a month so I can provide a monthly snapshot to my peers? 





  • Jessie Schutz
    Zendesk team member

    Hi Lettie!

    I've pinged our Community Moderators to see if any of our Insights experts can help you out!

  • Amy Dee

    Hi Lettie! At first glance, there are a few things I would check for a report like this.

    First, it helps to clearly define the relationships in a metric like this. It looks like you're trying to count individual tickets with reply times under/over 48 hours. I recommend this type of construction:

    • SELECT # Tickets WHERE (SELECT First Reply Time (hrs) [Mdn] BY Ticket Id) <= 48
    • SELECT # Tickets WHERE (SELECT First Reply Time (hrs) [Mdn] BY Ticket Id) > 48

    # Tickets and First Reply Time (hrs) [Mdn] are both under Metrics, while Ticket Id is under Attributes.

    This construction means Insights will check the first reply time on each individual ticket, then count the ticket based on whether its results are in the right range. I've always found this approach to be more reliable in a larger report.

    Second, I'd only use the = sign on one of the two metrics (like in my example above). That will keep tickets with exactly 48 hours from counting in both metrics.

    Third, check the filters on the rest of your report. In particular, this type of metric will respond to ticket dates and ignore event dates. If you're using a Date (Event) filter, the metric will ignore it and show you all tickets for all time. That could skew your results.

    I hope this helps! Happy reporting!

  • Steve Mooha

    Hey, I am looking for a way to create a report based on SLA and this got me close, but is missing something critical. 

    Tracking SLA time in a table. 

    # Tickets where SLA is >10Hours
    # Tickets where SLA is >0 Hours, < 10 Hours
    # Tickets where SLA is > -10 Hours, <0 Hours


    Is this possible? I am currently thinking about using time since Next Response as mentioned here:

    Anyone have any ideas how to implement something like this?

  • Steve Mooha

    As an update, I was able to get it working with the amount of days out of SLA through the (Event) how & filter. But I cant see anyway to do hours :O 

    Does anyone have any ideas? 

  • Sergei

    Hi Steve,


    From your question it seems like "Step:1 Creating custom metrics" from the above article is the answer. Worth mentioning, that calculations are done in minutes, so for your specific needs you would use 600 minute values. After building your custom metric (it will become an attribute that can be used later), it can be added to your report, and that will allow you to filter by hours. 


    If you are still stuck with this problem, it's worth dropping us an email to and explaining where exactly you are stuck. 




  • Steve Mooha

    Hi Sergel!


    I very may well drop an email, because I am not 100% sure this relates, but I would love to be wrong ^^ Just for awareness, its my understanding that FRT only tracks the time related to the very first agent response, and not the ones after that.

    We want our dashboard to show all tickets, for example, over an hour without a response, regardless of that being that ticket being touched one, two, ten or even fifty times!

    I will admit though, that I have not experimented with the custom FRT metric, and its now high on my TO-DO list :) 

  • Katarina Ondrejovicova


    Is there a way to actually get the list of tickets that belong to one of the categories? Lets say I want to get the list of tickets that are counted under the "more than 24 hours" portion of the chart.


    Thank you

  • Brett Bowser
    Zendesk Community Team

    Hi Katarina,

    You should be able to drill into the metric using Ticket ID. More information in our Adding drill-in to your Insights reports article.

    Hope this helps!

  • Sam Ashton

    Hi there, one of our channels has a 1st response time and second reply time under the SLA policy for that channel. Is there any way we can report specifically on the 1st response time when the policy is reporting for both?

  • Amy Dee

    Hi Sam! That should be possible. The recipe in the article above is based on the native first reply time metric. It doesn't use your SLA policies at all, so it may work for your needs.

    If you do start working with SLA reporting, you can use the SLA Policy and SLA Metric attributes to specify which policy and target you want to measure. That should allow you to exclude next reply time targets from your results.

    You can review the SLA attributes in our Insights object reference, and you can review the default SLA metrics in our Insights metrics reference.

    I hope this helps! Happy reporting!



Please sign in to leave a comment.

Powered by Zendesk