Even if you ultimately reject an argument, it may be that you were antecedently disposed to have greater credence in the premises (conjoined) than in the conclusion. Upon having this mild inconsistency pointed out to you by the valid argument, you will presumably revise your credal state by some appropriate combination of reducing your credence in the premises and raising your degree of belief in the conclusion.
If an argument can, in this way, be expected to lead many to rationally revise upwards their degree of belief in its conclusion, then it ipso facto has rational force (and is not question begging), in relation to this community of inquirers. This may be so even if their resulting credal state still falls short of full-blown belief.
I note that this approach also helps to refute a similar claim I addressed last year, namely:
No rational argument then can be constructed to ever change a person’s mind, because we can never get to premises that people must accept.
My earlier points refute it well enough, of course, but 'credence' talk makes it even clearer. For this reminds us that we may change a person's mind even without changing their full-blown beliefs (though we may do that too). Even if we merely sway them from thinking a view wildly implausible to instead thinking it moderately unlikely, that is clearly dialectical progress.