Two dogs can be taught to sit when a stranger approaches, and only one of them will be comfortable doing it. The difference is what the training addressed.
Behaviour and emotion are separable targets
Most training works on behaviour: an action is rewarded and becomes more likely. The dog's feeling about the situation is not directly involved.
Counterconditioning works on the association itself. A trigger that currently predicts something unpleasant is repeatedly followed by something good until the prediction changes.
The dog is not being asked to do anything. It is being shown a new relationship between two events, and the behaviour changes as a consequence rather than as the target.
Why order matters more than anything else
The trigger must come first and the good thing second, without exception. This is the detail that most often goes wrong.
Producing food before a stranger appears teaches that food predicts strangers, which makes food a warning signal rather than making strangers safe.
The sequence is trigger, then reward, every time, with a short and consistent gap. Reversing it can worsen the problem it was meant to solve.
Why suppressed behaviour is not resolved behaviour
A dog punished for growling frequently stops growling. The underlying discomfort is unchanged, and the warning that preceded escalation has been removed.
The result is a dog that appears calmer and is in fact less predictable, because a stage in its communication has been trained away.
This is the practical argument for working on emotion rather than appearance. Behaviour that has been suppressed reappears under pressure; a changed association does not.
How progress is measured
The useful indicator is anticipation. Early on, a dog sees a trigger and reacts. Later, it sees the trigger and turns towards the owner expecting the reward.
That turn is the observable moment the association has flipped. It is a more reliable measure than the absence of barking, which can simply mean the dog has shut down.
A dog that is still, silent, refusing food and holding a fixed posture is not progressing. Stillness and calm look similar and mean opposite things.
The pace it requires
The work is slow because it depends on repetition below threshold, and each session above threshold sets it back further than a session below it moves it forward.
Sessions are short, frequent and deliberately easy. Ending while the dog is still comfortable is more productive than pushing for one more repetition.
Where the trigger is unavoidable in daily life, management alongside the training is essential, and cases involving fear of people or a bite history need professional supervision.