There is a concept in moderation called ‘moderation based on intent’.
A three-strike system or similar might be general policy, but it isn’t always the optimal approach, and sometimes further nuance is necessary.
For example, a newly created account begins to spew bigotry. Should moderators wait for the third occurrence to take action, or should they just be banned immediately? They should be banned immediately, as all parties are aware that there is no intent for positive contribution here, and allowing the disruptive entity to remain causes unnecessary distress to other users.
Offering a path to redemption via a strike system is futile and a waste of resources in such cases. It can also be weaponised and exploited.
Intent can also apply to other situations and be gleaned from context. For example, the term ‘faggot’ could be offensive if used as a slur but benign if used to refer to a bundle of items. The appropriate consequence, if any, is determined by the intent behind the usage of the word.
A ‘based on intent’ clause has been policy for every stable community I’ve moderated, from small groups to large platforms.