tderz It makes sense to explain the order of moves with a real example. Here is a sequence how this shape occurs frequently in games:
Black wants to make White overconcentrated on the right side, e.g. there might be a small white wall (too) nearby.
(Only) In that case it does not matter if White is made even stronger by the black sacrifice - and - any chance of attacking White - is lost.
Of course White might try and resist otherwise.
Of course, White must capture the black stone . However, before doing that he still can play two useful kikashi - and - only in the right order. Kikashi is all about timing. This is an easy problem and kikashi/timing problems belong to the most difficult problems in Go.
after the - exchange would not work, because (at ) would not be atari anymore.
The wall is now much too close to , .
That this is not only an academic exercise is shown by the aji around a, after which White could threaten to put the stone in motion (life) or - after Black defends around b, do something else useful with the stones around a.