I am thinking of Digg here which is a good example of a redesign that led to a significant outflow of users.
Digg
For a while it was looking like Voat might be the new end destination after Reddit, but that was political rather than aesthetic.
It's not that I have signed any petitions, it's that with new feature set and presentation, there wasn't any place in my life for this. I still scrobble but I'm not really visiting.
That, and failure to bring in positive change. Instead of addressing problems that I am seeing (e.g. trolls and bitter people spamming broadcasts of artists) they're tweaking colors. Ok, bye.
What would you like to see in terms of significant new features and changes?
I used events (to show that I'm going to a live concert).
I had friends there. Another thing is people stopped using last.fm eventully.
I did artsts tagging and used shoutbox. This is where lack of community tools really kicked in. Many shoutboxes were outright dominated by undesirables.
- "Different" always takes getting used to. One's central neural network literally has to rewire new connections (and/or change the impact of existing ones) - for things that one uses often enough to have a dedicated circuit (which does not mean those neurons that are involved doing _only_ one thing, those exist [0], but more often it's a network thing). Heavy daily users of a service probably do have circuitry responding to that service in particular, and it has to reorganize itself a bit after the service changes. However, "different" does not have any bias towards a direction, "better" or "worse". It is just resistance to change because change is costly.
- Whether the change actually _is_ for the worse (difficulty: define objective criteria).
The initial gut reaction alone cannot be used to differentiate between the two scenarios. The additional difference is that the first point always happens and can be traced to real effects, but the second one can be subjective, you first have to define what the overall purpose of something is in the first place, which in the case of some non-(life-)essential website is relative (to whom you ask) and also not static.
Even if a change is for the better by some objective criteria, the effect of the resistance to the change itself will always be there. So initial reactions that are non-specific and only are about "the change is bad" are not all that useful. The magnitude of the reaction also may be a symptom of either the popularity (if people spend more time with something any change will impact them harder), that something really is worse, or a mix of both, hard to tell on its own.
[0] https://www.scientificamerican.com/article/one-face-one-neur...