About two and a half years ago, we released RKG Duck, an open-source tool that lets you use a Perl filter on the Windows Clipboard.

We previously discussed how one could use this tool to create AdGroups by filtering a list of keywords to cluster similar phrases together using a common signature. For each phrase, our filter removed stop words, took the stem of each word, alphabetized them, and crushed the stems together into a signature.

Today, we would like to introduce the RKG RegExp Duck Filter as an alternative method for creating and organizing AdGroups, using the words within the phrases themselves. Here’s what the filter looks like:

# regexp filter
# input from STDIN, output to STDOUT

use strict;

# Enter your word list pattern on the line below
# Example: …= “red|sony|camcorder”;
my $regexp = “<INSERT HERE>“;

{
my $i = 0;
my %sort_order = map {ucfirst lc $_ => $i++} split(/\|/, $regexp);

sub list_order {$sort_order{$a} <=> $sort_order{$b}};
}

while (<STDIN>) {
print join(”, sort list_order map {ucfirst lc} m/\b($regexp)\b/gi), “\n”;
}

# RKGDUCK
# documentation at www.rimmkaufman.com/duck

By modifying the filter in RKG Duck and defining the “$regexp” variable as a pipe-delimited regular expression (ex. “red|sony|camcorder”), any phrase containing “red” and/or “sony” and/or “camcorder” will be categorized as some variation of “RedSonyCamcorder” (ex. “RedCamcorder,” “SonyCamcorder,” etc.)

Additional descriptive words can be used in the regular expression but should be listed and grouped together to maintain naming consistency (ex. color-brand-product-accessory: “blue|red|sony|samsung|camera|camcorder|lens” instead of “blue|samsung|camera|lens |red|sony|camcorder” so that a blue Sony camera lens is categorized as “BlueSonyCameraLens” instead of “BlueCameraLensSony.”

Check out this sample spreadsheet to get a better idea of what you can expect and how to use the filter. Then, download ActiveState’s ActivePerl, RKG Duck and the Regexp filter today to try it out for yourself!

Note: Change the file extension of the filter to “.pl” before using it with RKG Duck.

This example is a demonstration of what one can do with RKG Duck. This isn’t how our firm builds AdGroups — we don’t recommend building out phrases then AdGrouping later. Rather, the right approach to build campaigns is to go through every relevant URL on the site; build out rich keywords for each URL; assign targeted copy, smart match types, and sensible initial bids; then adgroup them appropriately. As George mentioned, there’s significant benefit in databasing ads in a richer taxonomy than the simple Campaign >> AdGroup >> Term hierarchy. However, if you’ve taken over a badly managed program, this can be a useful way to right the ship fairly quickly.

Follow these links to learn more about regular expressions or CPAN modules to create your very own RKG Duck Filter!