Update KB helpful survey #6423
Update KB helpful survey #6423
Conversation
There was a problem hiding this comment.
The use of htmx and alpine.js is pretty cool!
-
The GA4 tracking events are not being generated (see comments in code below).
-
When the user is not authenticated but has already voted anonymously, the voting form is still displayed when the article is viewed again, and if another vote is attempted, the response from the vote POST is a redirect to the KB article, which causes a strange article-within-an-article view of the page.
- The font size and weight of the
YesandNoof the voting form should match those of itsWas this article helpful?text.
- When the user is authenticated, the voting form is pushed way down on the page.
- The
CancelandSubmitbuttons of both the "yes" and "no" surveys are stacked instead of side-by-side.
- When submitting the survey, the response is
Thanks for making us better!instead ofThanks for your feedback!as the Figma indicates. Also, the text is not bold as is shown in Figma.
I am not sure about this. We treat the
It's actually the same space with the logged out view. The difference is that the editing tools push down the whole section
Have you tried a fresh build? It seems to be working for me |
akatsoulas
force-pushed
the
update-kb-survey
branch
from
12cfa7b to
b7b3117
Compare
Ah, I see that now, thanks. Ignore that comment! 😄 Regarding the cancel and submit buttons of the survey, I did have a fresh build, but they're still always stacked for me. However, I noticed that your survey is wider than mine, so I guess the width of my survey is what is causing the buttons to stack. Hm, I'm not sure why my survey has a smaller width. Anyway, ignore that one too. 
|
I will have another look to make sure that it scales properly to different screen sizes |
There was a problem hiding this comment.
Overall, I really like the GA4-event-related changes, and when I tested them locally, the GA4 events are generated properly (except for the first comment below). I have two comments:
- After submitting a survey response, a final pair of
survey_openedandsurvey_closedGA4 events are generated, becausesurvey_form.htmlis rendered and returned for theresponse_messageofThanks for your feedback!. Both of those survey events include an emptysurvey_typeparameter. See the image below. I'm thinking we shouldn't generate survey events in this case, because it'll misleadingly inflate the GA4 survey events.
- For the survey events, maybe we should use
article_survey_...instead ofsurvey_...to distinguish these events from any possible future survey events?
| <button type="submit" class="btn helpful-button" name="helpful" | ||
| data-event-name="article_vote" | ||
| data-event-parameters='{"vote": "helpful"}' | ||
| >{{ _('Yes') }}👍</button> | ||
| <button type="submit" class="btn not-helpful-button" name="not-helpful" | ||
| data-event-name="article_vote" | ||
| data-event-parameters='{"vote": "not-helpful"}' | ||
| >{{ _('No') }}👎</button> |
akatsoulas
force-pushed
the
update-kb-survey
branch
from
b7b3117 to
0ab27e8
Compare
There was a problem hiding this comment.
The GA4 events look good now. Just an indentation nit, and there's still the issue of the text of the Yes👍 and No👎 buttons looking different than the Figma designs (but as you said, you can discuss that with Josh). Thanks @akatsoulas!
| const surveyType = '{{ survey_type }}'.trim(); | ||
| if (surveyType) { | ||
| trackEvent('article_survey_opened', { survey_type: surveyType }); | ||
| } |
There was a problem hiding this comment.
Nit. The indentation is off.
akatsoulas
force-pushed
the
update-kb-survey
branch
from
0ab27e8 to
92bf480
Compare
No description provided.