Abigail's solution is good. I have modified your code slightly to build a hash suggested by Abigail. Instead of hardcoding the hash assignment, I have created a list called col_def that holds the name of the data columns in your data, and use it to construct the hash entries dynamically.

Note that I have added use strict and use Data::Dumper to the top of the code. Data::Dumper is one of the most handy development utility while coding hash and array structures, it dumps the data structure held in the variable. It's a good practise to include both use strict and use Data::Dumper in your code, especially when you are learning to use Perl.

#!/usr/bin/perl -w
use strict;
use Data::Dumper;

my @col_def = qw/ id name street city zip phone start_dt /;
my %ACCTS;

open(IN, "< myfile") or die "Help!\n";
while (<IN>) {
    chomp;  # need to remove \n character, otherwise last
            # field will contain \n
    my @cols = split /\|\|/;

    $ACCTS{$cols[0]} = {
        map { $col_def[$_] => $cols[$_] } 1..$#col_def
    };
}
close IN;

print Dumper(\%ACCTS);
[download]