Populating Problem Ticket Subject in Explore
I'd like to show some queries looking at problem tickets and how wide spread they are, when/if they were solved, etc. However, I'd like to identify the problem by subject rather than ID. How can I create this custom metric?
I was able to get the subject to populate with a custom attribute:
IF ([Ticket type - Unsorted] = "Problem") THEN [Ticket subject]
However, I can't get this attribute and the Problem ID to populate at the same time
Hi Abigail Nichols.
I can see your formula and as you have enabled account assumption for us, I assumed into your account to check.
It is a great formula, and it looks right for me.
What in my opinion could be changed are the Metric (I see that COUNT(Tickets) would make much more sense), and instead of creating an Attribute "Problem ticket ID" simply add the ticket ID attribute to rows and filter by ticket Type>Problem.
Please see some screenshots from my test account where I recreated the attribute Problem Name Attribute:
As you can see, I also added a time attribute to filter and reduce the amount of data.
Hope it helps to clarify and feel free to send us any additional questions.
I need to be able to create a report with the # of incidents tied to the Problem Ticket Subject.
Because of this, using COUNT(Tickets) doesn't work, because it will only pull in the count of the problem ticket. Using Count(incidents) doesn't allow me to pull the problem subject, only the ID.
We need this to be able to track the feature requests we receive, and understand the number of customer requests we're receiving tied to each. We use the Problem Subject ID to provide the detail of the request. Sure, we can setup fields to track related by product and we do, but we need specific details about the request that we include in the subject line and not just the product its related to.
Is there anyway at all to associate the number of incidents w/ the problem subject line?
This shouldn't require assuming into our account, as it can easily be replicated with any setup in a test Explore environment. It is not specific to our setup, and we cannot provide access due to protected health information within our Zendesk.
It should be simple enough to recreate this in a test environment of your own. All you need is a Problem ticket with some linked incidents. We need to be able to find a way to report the incident count next to the Problem ticket Subject.
Please let me know if this is possible.
Can I expect an answer on this anytime soon?
We're trying to do the same thing. Is there any way to sub in the name for the ID?
Same issue- trying to put together a list of problems with number of incidents, but want to display the subject of the problem, not the ID number.
Hello! I figured out a solution to this problem. The key is to set the metric to Problems, and then use your create metric to get the metric subject name. Here is a screenshot of it working as well as the custom subject name. If you want to get incidents you will need to do a separate table all together.
Hi @Kelly-Anne Richardson
I appreciate the assistance, but that is not a solution to this issue.
You don't actually need a custom attribute at all if you want to get the Problem ticket subject name - you can just filter by ticket type and use the regular Subject attribute when using the regular Count(tickets) metric.
The primary issue here is that Explore has disconnected Problems from their incidents - mainly, because the only thing that ties an incident to a problem is the Linked Problem Ticket field, and while that is reportable, no details related to the problem are returned in any scenario -- all details will be tied to the incident itself.
I just wanted to call this out because it is not a solution, and frankly Zendesk needs to respond to this issue as it's a significant oversight that prevents seamless reporting on problems and incidents.
Hi Kelly-Anne- Thanks so much for jumping in!
Unfortunately, Kate is right. We're trying to show the name of the problem ticket *that's tied to an incident*. Or, more specifically:
1) The problem ticket subject and then the number of incidents tied to it, or;
2) The problem ticket subject and then a list of incidents linked to it.
A HUGE plus would be clickable links... but that may be too much to ask. ;)
Ah I see, sorry I read that initial problem wrong. I do have a "solution" to this but its INCREDIBLY hackey and what Zendesk Support told me to do. However, I do want to post it regardless for someone else who is reading this and needs this data TODAY for whatever reason. I absolutely agree Zendesk needs to address it and create a better solution because frankly its' rather silly I have to do this. This will only work if you are using this data outside of Zendesk Explore (unless there is a way to merge tables after they are already created that I am unaware of). For example I was using this data in a power point presentation.
Here is the hack: I created two separate tables and then externally (after an export) in excel used xlookup to merge the tables on the ticket ids.
Here is a screenshot of how to make the first table.
here is a screenshot of how to make the second table.
Hi @Kelly-Anne Richardson
Thank you for the additional information.
On your first screenshot, is that just an image inserted from your excel table?
I have historically been simply exporting and doing a similar method of generating the tables in excel as needed. However, this has caused issues as essentially I have to re-export on some regular basis since it won't auto-update with new incidents/new problems.
This places a huge hurdle in the way of seamless and efficient reporting on bugs and feature requests with our Product and Engineering teams)
I still can't help but feel that Zendesk is really dropping the ball on this one - I know they don't even use Explore internally (data is in Tableau) and it feels like they've just abandoned any improvements to their own tool.
The first image is of the two tables in Zendesk Explore I created. I underlined how Problem Ticket ID and Ticket ID were returning the same values. And I can confirm if you look up that ticket number in support those incident counts from the second table and the subject line from the first table do in fact match.
Let me clarify - so, you are manually filtering by ticket problem ID then to build the tables?
Hi, sorry I am little confused by your question. I am just going to explain how to use xlookup for anyone else who might be reading this. I apologize if this is stuff you already know.
The first table returns Problem Ticket ID, Date Created, Ticket Status, and Ticket Brand, and Problem Ticket Subject
The second table return Ticket Id, and incident counts.
I export them BOTH tables into excel. Lets say I put them into one page in excel for the sake explaining and it looks like this.
Then I use xlookup which is an excel function you can think of as a merge or join. I highlighted the part I "do by hand" in yellow. When I say do by hand I mean I write one formula and then apply it to the rest of the column.
Its looking at the rows ticket id, and referencing the Ticket Problem Id and looking for matches. If it matches it returns the number of incidents of that row and it looks like this after you click enter after finishing the formula.
That formula put that 70 there, I did not do that by hand. Then I just apply that formula to the rest of that yellow column. If you have these tables on different excel worksheets, you can literally just drag the formula down the column, no need to do anything by hand and it auto fills in everything.
Thank you again for the detailed explanation - I definitely appreciate your effort!
Apologies for not making myself clearer - the excel portion makes sense, but my question is how you got this back into Explore without either displaying all problem tickets or manually filtering the incident report on Problem TIcket ID.
After doing the semi-manual work of creating the xlookup formula/table, how then did you get this to populate in Explore without manually entering the Problem IDs into the Ticket Problem ID?
Perhaps if I provide an overview of what I'm hoping to achieve it will help you understand my dilemma:
- We track both bugs and Feature Requests as Problem Tickets
- We are hoping to create a single report which displays the Problem Ticket Subject along with it's incident counts (so essentially it would be a table with columns Problem Ticket ID, Problem Subject, # Incidents)
- We're currently limited that we can get one table with Problem Ticket ID and Subject (call this chart A), and a second table that shows Problem Ticket ID & # Incidents (call this Chart B)
- What this requires then is one of two things: we either show all problem tickets on both reports and individuals have to map the IDs between the table with the subject and the table with the incident counts, or we have to manually filter on Problem Ticket ID to only show the subset displayed in the Problem Subject Chart
- If Chart A is filtered to only show a subset of problem tickets (as we filter out each by the ENG/Product team responsible), then Chart B needs to be manually filtered on Ticket Problem ID to match what's returned in Chart A (I'm aware we could probably get around this manually filtering with additional ticket fields, etc. but that is not an elegant solution and I would prefer not to follow along)
At this point, I know I can export this and create a chart in Excel, but that data is static and will not update when new incidents are added, or when new problem tickets are created, which means that I have to setup a regular cadence of exporting/recreating the chart in Excel/sharing this out <-- this is what I currently do, and it takes a lot of time.
What I need is a means for Explore to connect Problem Ticket Subjects w/ Incident counts - pretty basic, and I shouldn't have to go through a million steps to get a report that accurately updates how many incidents are linked to a Problem ticket w/ its subject. I can think of other manual ways to capture this (heck, I can add a numeric field and make my agents update the problem to +1 the vlaue in that field every time a new incident is added, but this again is not ideal, people will get busy and forget, and it shouldn't be the workaround given that Zendesk clearly knows how many incidents there are, just they don't surface that anywhere as an attribute of the Problem ticket for reporting purposes)
In the screenshot you showed from your first comment - it looks like you're essentially doing the same thing as I am now: creating one report which shows the Problem Ticket ID & the Subject, then the second chart which has the Ticket ID (in this case, the problem ticket) & the incidents count and its up to the viewer of the dashboard to tie the two charts together. This is what I'm hoping to avoid.
So essentially my question is -- it looks like, in the dashboard in Explore, you're doing one of two things:
- Displaying all problem tickets on both reports (not a solution we can follow, as again we need to filter the problems & incidents by subset of who is receiving them)
- Manually filtering on Problem Ticket ID on the incident count report with the Ticket IDs from the Problem Subject report (not a solution, as it requires manual intervention)
Are you doing something other than one of the two above?
Ah this is where we differ, in our projects.
My project is a once a month report where they want all of the data in a Powerpoint. So I am using explore to get my data, but it's not my final output. So using excel works for me since I am pasting the final table into a power point, not explore. Yes, what I did above is all I've done and it works for me because of the scope of my project. Sorry I can't help more to you. I hope I can help someone else who may be reading this thread though.
No worries - I genuinely appreciate the information you've provided so far, and appreciate you taking the time to share so much of it!
Hello Admins and Explore analysts,
I realised long ago that this is simply not possible using the current state of Explore, so I now use python to automate my reports to csv. I published a sample that you may find useful if you are familiar with basic coding: https://github.com/ahmedsabriz/zenpy-incidents-report
python3 incidents_report.py --start "2022-01-01" --end "2022-05-15"
Creates a csv of the number of incident tickets created between those dates by problemID AND problem subject:
Any updates from Zendesk on this issue? Why do they have a feature in the application that we cannot report properly on?
We're sorry if this is not yet possible in Explore. If you have time, I encourage you to create a new post in the Explore Product Feedback topic in our community to engage with other users who have similar needs and discuss possible workarounds. Conversations with a high level of engagement ultimately get flagged for product managers to review when they go through roadmap planning. Thank you!
For anyone following along here, I just created this Feedback topic related to this Q&A (per Gab's suggestion), you can see it (and upvote it, if you'd like) here: https://support.zendesk.com/hc/en-us/community/posts/5448109683994-Populating-Problem-Ticket-Subject-in-Explore-
Please sign in to leave a comment.