pontoon icon indicating copy to clipboard operation
pontoon copied to clipboard

Add specs for Email consent opt-in and Unsubscribe page

Open bcolsson opened this issue 1 year ago • 3 comments

Fixes #3109

bcolsson avatar Feb 21 '24 22:02 bcolsson

Not sure if this is ready for review yet, so just adding an implementation note before I forget about it: We should make the consent text configurable per instance.

mathjazz avatar Feb 22 '24 08:02 mathjazz

As agreed with Bryan, flagging @flodolo and myself for the final review.

mathjazz avatar Feb 26 '24 17:02 mathjazz

@flodolo - I've added a draft of the unsubscribe page content.

bcolsson avatar Feb 29 '24 00:02 bcolsson

I'm happy to take a stab at that and work on the missing mockups (we should have one at least for the Unsubscribe page).

@mathjazz - Thanks, that would be greatly appreciate.

bcolsson avatar Mar 11 '24 22:03 bcolsson

Depending on the opt-in rate we might not be able to reach as many users as we'd like. How do we mitigate this risk?

Should we — after a period of time — also show the opt-in form to users that have declined to opt-in in the past?

mathjazz avatar Mar 12 '24 06:03 mathjazz

Depending on the opt-in rate we might not be able to reach as many users as we'd like. How do we mitigate this risk?

Should we — after a period of time — also show the opt-in form to users that have declined to opt-in in the past?

I'm not sure if we'd want to do this too often to respect the user's choice, otherwise we risk looking like a dark pattern (nagging).

In my research on opt-ins, I've read that it's considered best practice to try and regain permission from users who have not interacted with your emails after 6 months. Since we wouldn't have the ability to track that, we could instead show all users at 6 month intervals a screen to confirm their email communication preferences. What do you think about that?

bcolsson avatar Mar 13 '24 22:03 bcolsson

Depending on the opt-in rate we might not be able to reach as many users as we'd like. How do we mitigate this risk? Should we — after a period of time — also show the opt-in form to users that have declined to opt-in in the past?

I'm not sure if we'd want to do this too often to respect the user's choice, otherwise we risk looking like a dark pattern (nagging).

In my research on opt-ins, I've read that it's considered best practice to try and regain permission from users who have not interacted with your emails after 6 months. Since we wouldn't have the ability to track that, we could instead show all users at 6 month intervals a screen to confirm their email communication preferences. What do you think about that?

Following-up off-line with @flodolo, I think in this case we should respect the user's choice at time of setting the preference. To promote additional opt-in and mitigate against this risk we should use our channels (socials, events, etc.) to remind people to sign-up / change their preferences.

bcolsson avatar Mar 18 '24 21:03 bcolsson

Here's my proposal for the designs of the standalone opt-in consent pages. I tried to keep it simple and concise, as that's a general good practice to increase conversion rate. It's also consistent with most of the email subscription forms I've found online, including the newsletter subscription forms on Mozilla websites: https://blog.mozilla.org/#newsletter-content https://www.mozilla.org/

All the designs are the same, just the title is different.

I wonder if "By enabling email updates, I'm okay..." should in fact be an off-by-default checkbox. Legal will know.

1 2 3 4

mathjazz avatar Mar 20 '24 20:03 mathjazz

I'm not sure if we'd want to do this too often to respect the user's choice, otherwise we risk looking like a dark pattern (nagging).

In my research on opt-ins, I've read that it's considered best practice to try and regain permission from users who have not interacted with your emails after 6 months. Since we wouldn't have the ability to track that, we could instead show all users at 6 month intervals a screen to confirm their email communication preferences. What do you think about that?

Absolutely - we shouldn't do this more often than once or twice per year. We'll need to track when was the consent given or not for each user, so we should add that to the spec.

External channels are typically much less effective in steering users to take an action within the platform, so I'd consider that as a bonus.

mathjazz avatar Mar 20 '24 20:03 mathjazz

@mathjazz - I've updated copy and mockups with suggestions from copy editing. I've also unified the consent ask page to be the same for existing and new users. Unless you have objections, I'm going to present the spec we have here to legal.

I will also be asking their opinion on whether or not asking for consent again (regarding this open question).

bcolsson avatar Apr 04 '24 19:04 bcolsson

No objections, looks great!

mathjazz avatar Apr 04 '24 19:04 mathjazz

I've updated the spec with all of the legal feedback.

Regarding the one outstanding question of whether we want to show a reminder to users about opt-ins: there weren't any legal concerns if this happened in-app. (Sending an email about this to users would not be okay however.) So this option remains open to us. However, I haven't added this to this spec document to limit the scope to what we need to start gathering permissions now.

I have added the requirement that we note that we record the user's choice of opt-in or not, and the date and time they made the choice. This way we can add it in a future update.

bcolsson avatar Apr 11 '24 23:04 bcolsson