We assume that some nontrivial portion of users will choose to use the widget by bookmarking the page. For that portion of users, some will find it annoying that they must take action to start the widget -- by clicking the toggle -- every time they load the page.
@oxtoacart is a concrete example of this hypothetical kind of user. On Slack, he said:
After I've gone through the onboarding once, there's no scenario in which I would choose to open the widget page and not choose to toggle it on, and there's also no scenario in which I would toggle it off but leave the page open (except by accident).
So from my perspective, the toggle is just noise/overhead.
A lively Slack conversation produced two possible approaches to improving this.
From @myleshorton:
You could have a check box "always turn on when page loads" or something, but this also reminds me of lantern proper where you still have to turn the thing on when you open it. That seems like a relatively minor point of friction honestly, in exchange for more clearly putting the user in the drivers seat
From me:
I may be hallucinating this, but did Snowflake have a feature where it memorized your toggle setting with a cookie and automatically set it to "on" when you came back to it?
Worth mentioning: It's entirely possible that our initial assumption is a bad one, and that in practice, nobody really uses the widget by bookmarking the page, so none of this really matters. We may discover that the page serves only as a gateway for users to get introduced to the project, and that nearly everyone uses the widget via the extension or an embed.
We assume that some nontrivial portion of users will choose to use the widget by bookmarking the page. For that portion of users, some will find it annoying that they must take action to start the widget -- by clicking the toggle -- every time they load the page.
@oxtoacart is a concrete example of this hypothetical kind of user. On Slack, he said:
A lively Slack conversation produced two possible approaches to improving this.
From @myleshorton:
From me:
Worth mentioning: It's entirely possible that our initial assumption is a bad one, and that in practice, nobody really uses the widget by bookmarking the page, so none of this really matters. We may discover that the page serves only as a gateway for users to get introduced to the project, and that nearly everyone uses the widget via the extension or an embed.