Sign up analytics#767
Conversation
davinotdavid
force-pushed
the
features/164-sign-up-analytics
branch
from
207898b to
3fedf67
Compare
| const nextStep = () => { | ||
| let nextStepValue = SignUpSteps.INVALID; | ||
| let nextStepValue = SIGN_UP_STEPS.INVALID; | ||
| switch (step.value) { | ||
| case SignUpSteps.USERNAME: | ||
| nextStepValue = SignUpSteps.PASSWORD; | ||
| case SIGN_UP_STEPS.USERNAME: | ||
| nextStepValue = SIGN_UP_STEPS.PASSWORD; | ||
| break; | ||
| case SignUpSteps.PASSWORD: | ||
| nextStepValue = SignUpSteps.VERIFY; | ||
| case SIGN_UP_STEPS.PASSWORD: | ||
| nextStepValue = SIGN_UP_STEPS.VERIFY; | ||
| break; | ||
| case SignUpSteps.VERIFY: | ||
| case SIGN_UP_STEPS.VERIFY: | ||
| return; | ||
| } | ||
| captureStep(nextStepValue); | ||
| step.value = nextStepValue; | ||
| } |
There was a problem hiding this comment.
If we are only doing the captureStep when the user moves to the nextStep, we are never capturing the initial SIGN_UP_STEPS.USERNAME.
So maybe we should add a captureStep() call in the onMounted of the Step1 component here:
There was a problem hiding this comment.
davinotdavid
force-pushed
the
features/164-sign-up-analytics
branch
from
2e3a27f to
d90e919
Compare
|
After discussing at the last thundermail meeting, we're good to go with these changes. |
I missed that unfortunately, how are we allowing the anonymous user to opt-in? Or was it auto opt-in? |
I defer to @Sancus (who articulated this during the last meeting), but it's auto opt-in. Users are invited to this early/experimental phase, during which we're figuring out if the FTUE has any serious issues with usability and functionality. Instead of individuals having to fill out a support form, we're automatically checking for significant drop offs during the sign up flow. |
|
okie dokie. Just checking because last I heard @kewisch had some issues here. |
|
Cool! I will rebase this soon™ (today, in a few) |
|
@MelissaAutumn sorry you missed this, we talked about it in the last accounts meeting. We're going to do a time limited experiment to see if what we're collecting here will actually provide benefit and is answering the questions we're looking at. All things considered, as long as we're not collecting PII and determining only in aggregate where people are leaving the sign up process, we're not being overly invasive. When we consider more elaborate data points or are considering this a more long term experiment we should definitely have some form of user choice, though we'll need a bit more input from legal counsel to do this right while not artificially limiting ourselves or creating bias in the data that we receive. |
|
Sounds good to me, as long as it's all clear and no PII is stored. |
davinotdavid
force-pushed
the
features/164-sign-up-analytics
branch
from
d90e919 to
803007f
Compare
|
@radishmouse @MelissaAutumn Alright, I think we are good to go but since I helped with a commit, rebased and reviewed it might be good to do one final review by someone else. Feel free to merge it in if so! |
|
Looks okay to me, thanks folks! |
4 participants
Fixes #164
We don't have a support link, but I'm sure we'll find a spot to stick that in eventually.
Eventually I'll rework the folder structure and setup actual .env files that will control some defines, but for now it's a hardcoded true.