I applaud your willingness to post an actual example script and actual example data. Please allow me to offer a number of points:

• It's often better to specify a minimum version of perl. See also Perl::MinimumVersion.
• A small amount of sample data is often as good as more.
• Limit lines to 78 columns. Continuations are okay. If I really did need to hardcode a big paragraph, though, I might read it in from <DATA>. See perldata, Special Literals.
• Proper indentation and spacing go a long way toward making code readable.
• Use interpolating quotes only when you want to interpolate.
• atcroft is correct in that your original script attempts to perform substitution on $_ ("it"); if successful, the expression evaluates to the number of substitutions performed.

The following script attempts to illustrate each of the points above.

use 5.010000;
use strict;
use warnings;
use diagnostics;

my @foo     = (qw| Pear Apple AppleTomatoApple |);

for ( @foo ) {
    my $text    = 'Apple has determined, '
                . 'the iPod nano '
                . 'may overheat. '
                ;
    my $assign  = 'Blueberry jam vendor!';
    
    $text      =~ s/Apple/Mango/g;  # sub's on $text
    $assign     = s/Apple/Mango/g;  # sub's on $_ and yields num of su
+ccess
    
    say '$text: ', $text, "\t", '$assign: ', $assign, "\t", '$_: ', $_
+;
};
[download]