Andy F.

Rafael Santos ah of course! I was overthinking this, believing I had to add the exclusion to the JSON when of course I can just add it to the main trigger. Thank you!

Hi Zendesk community! I have a question about using Liquid Markup with Zendesk ticket tags. I've set up A/B testing in our instance (as described in this article: https://ro.roca.work/blog/ab-testing-in-zendesk) and so far, it's mostly been working for us. However, I just identified a scenario that I can't figure out how to solve.

To summarize, every ticket that is created gets assigned either the control or experiment tag. The problem occurs only sometimes with tickets created as followups, as it's possible for a ticket to carry over the previous tags, like control, but then get assigned the experiment tag by the webhook. How would I use liquid markup in this JSON response to tell it not to fire if a ticket already has the control or experiment tag?

{% assign randomizer = ticket.id | modulo:2 %}

{% case randomizer %}

{% when 0 %}

{“ticket”: {“additional_tags”:[“control”]}}

{% when 1 %}

{“ticket”: {“additional_tags”:[“experiment”]}}

{% endcase %}

Thanks in advance for the help!

Thank you Brandon! Really appreciate these suggestions!

Thanks for the suggestion, Jonathan! We've started messing around with a few subject line tests, but we haven't explored the possibility of adding more visuals yet. It's definitely worth exploring!

A question for the community - what A/B tests have people had success running on tickets in their instances? I was able to get A/B testing working (all tickets are marked as control or experiment on creation, and we can then run tests based on those tags) but I'm having trouble coming up with ideas on things that we can test. Main goals are improving CSAT and agent efficiency, but any and all ideas are welcome!

Hey Zendesk team, it doesn't look like there's an easy way to do A/B testing with tickets in production right now through Zendesk's native tooling. Currently, it looks like the only way to accomplish this is through some custom tooling of webhooks and APIs, as detailed in these posts:

https://support.zendesk.com/hc/en-us/community/posts/4409515169306-Using-Liquid-Markup-to-A-B-Test-Your-Triggers

https://roca.work/blog/ab-testing-in-zendesk

I'm personally learning about Webhooks and APIs as I can to try and skill up to do this, but it's a pretty significant barrier to entry that requires some very specific technical knowledge. As we're definitely not the first to try and accomplish this goal (as per the existence of these articles), a native A/B testing solution in Zendesk would be amazing. 

Boiled down to the simplest requirement, all that's really needed to make this feasible is a trigger or rule that causes all incoming tickets to be tagged as A or B (or A/B/C, A/B/C/D, depending on the test) and alternate between them / rotate through the list. Then with a test group (A) and a control (B), abracadabra, you have working A/B testing in Zendesk.

For a short term fix, would it be possible for an Engineer to implement trigger conditions / logic that achieve the described rotation above?

Hey Zendesk, just a heads up that this request has been open for over three years now, causing us to lose confidence that Zendesk will actually make enhancements based on user feedback. Is this typical, in that user requests don't get addressed for over three years?

+1 to all the above comments. This is now six years since the original request was filed, and all we want is the ability to change this to operate as per other tickets. In the old Facebook Private Messages, Enter did not send the message, so we all know that the functionality is there. Just give us the setting option, please.

+1

Adding my voice to the choir on this one, a back button that could be implemented in flow builder would be extremely helpful. Agree with Will and Masako.