use Benchmark;
my @letters = qw (A B C D E F);
my $string = "AFTYUBEWTWECRTUTYIYTDDDDRYJURTHJTREEEEEFGSDFF" x 100;

timethese(1000,{ match=>\&match,subst=>\&subst});

sub match{
    my $str = $string;
    foreach (@letters) {
        my $count = 0;
        $count ++ while $str =~ /$_/g;
    #   print "$_ : $count\n";
    }
}
sub subst{
    my $str = $string;
    foreach (@letters){
        my $count = $str =~ s/$_//g;
    #   print "$_ : $count\n";
    }
}

output:
Benchmark: timing 1000 iterations of match, subst...
     match: 10 wallclock secs ( 2.52 usr +  0.01 sys =  2.53 CPU) @ 395.26/s (n=1000)
     subst:  2 wallclock secs ( 1.88 usr +  0.01 sys =  1.89 CPU) @ 529.10/s (n=1000)