Good articles on confirm prompts vs undo

These address our patterns to Avoid mistakes and provide Undo. The text is
taken from the Smashing Newsletter.
2. Confirm Or Undo?

As designers, when it comes to critical actions in our UIs — like moving
records or archiving cards or deleting email lists — we tend to be cautious
and protective. We don’t want our customers to make costly mistakes just by
hitting a wrong button, or by tapping on a wrong spot unintentionally. So
we try to verify intent with confirmation dialogs, usually asking to
confirm the action, and sometimes asking to type in a word explicitly. But
is it always the right design decision?

[image: Confirm or Undo]
<https://smashingmagazine.us1.list-manage.com/track/click?u=16b832d9ad4b28edf261f34df&id=b277ccafea&e=3450dbc01d>

In his article, Confirm or Undo
<https://smashingmagazine.us1.list-manage.com/track/click?u=16b832d9ad4b28edf261f34df&id=a971274f4e&e=3450dbc01d>,
Josh Wayne argues that customers often don’t read confirm dialogs and
habitually click confirm. His suggestion is to avoid confirmation dialogs
and use *undo* instead: a visible notification allowing users to restore or
recover the initial state of things with a click. Kinneret Yifrah agrees as
well
<https://smashingmagazine.us1.list-manage.com/track/click?u=16b832d9ad4b28edf261f34df&id=f75ffa9e20&e=3450dbc01d>,
arguing that a decision will depend on reversibility of action, its
severity, complexity and frequency. The less severe the action, the less
likely confirmation dialogs to be needed.

As Alan Cooper writes, “For confirmation dialog boxes to work, they must
only appear when a user will almost definitely click the No or Cancel
button, and they should never appear when a user is likely to click the Yes
or OK button.” So perhaps we don’t need confirmation dialogs as much as we
think we do. *(vf)*