Content Ad Campaign Keyword Strategy Revisited

The keywords you choose for content ads play a different role than they do for search ads; a point that’s often tough for search advertisers to grasp. In traditional search advertising, keywords are intended to match search queries that are usually related to the advertiser’s products and services. The keywords in a content ad group should describe the types of pages and sites where you want your ads to appear.

The number one reason for content advertising campaign failure is because ads frequently appear on untargeted, irrelevant pages, but you can control ad placement more precisely by including a small number (20-40) of keywords in your keyword-targeted content ad groups.

In previous columns, I’ve described how content matching algorithms work, and explained the concept of themes — in Google’s case, a list of 594 known categories it uses to match content ad groups to corresponding AdSense publisher pages. I also suggested a method for building keyword lists that relied on intuition more than anything else.

Today, we’ll look at an alternative method that’s a little more scientific and should give you even more control over where your ads appear.

The Ultimate Keyword Tool…Does It Exist?

Let’s start with this assumption: if a keyword-targeted ad group’s keywords should describe the pages/sites where an ad should appear, then possibly the best keyword list is composed of words/phrases that appear most frequently on the target sites’ pages. Presenting. the thinking and evidence behind this hypothesis would take more time than we have here, so trust me for now, it’s a sound theory.

Armed with this realization, I looked for the ultimate tool for deriving such lists. Ideally, the tool would accept a list of URLs, load every word of content from all pages at the root and in subfolders of that URL, and return a ranked list of one- and two-word keywords.

I’m sorry to report that I didn’t find the ideal tool — but I found a few that do part of the job, and in combination, get close to perfection.

The first is Topicalizer, a free Web-based tool that accepts a URL (or a pasted list of words) and returns the 10 most frequently-appearing words, as well as the 10 most frequently-occurring two-word and three-word phrases. The problem? It only acts on a single page at a time, so deriving a consolidated list of many sites/pages would take way too much time.

Another tool is the Hermetic Word Frequency Counter Advanced Edition. This heavy-duty Windows software will analyze any number of documents stored in a local folder, and return a ranked list of the most frequently occurring words. It’s well worth the $48, but it falls short of perfection in one crucial aspect: it won’t return a list of frequently occurring phrases unless you specify the phrases to look for. Catch-22.

The tool that comes closest is the cryptically named Textanz. A bargain at $22.95, this Windows application takes any local text file (which can include any Web page file), and displays lists of the most frequently-occurring words, as well as lists of frequently-occurring phrases containing any number of words you designate. One main drawback: it will only operate on a single file, so it’s unable to process a folder representing every page of a site.

No doubt a vigilant reader will find a better solution — if that’s you, e-mail me, or start a discussion in the SEW Content Advertising forum. Hopefully that will happen before next week’s column, and I can describe it then.

Meanwhile, here’s a convoluted way to get close to nirvana:

  1. Use the Google AdWords Placement Tool to find a list of sites for a particular category/theme.
  2. Use a free tool like WinWSD Downloader to download each site to the local hard disk.
  3. Put all downloaded sites into a single folder and use the Hermetic tool described above to create a list of most-frequently-occurring words.
  4. Optional: use a tool like MonkeyMerge ($16) to merge all pages into one document, and then use Textanz to analyze the file for most-frequently-occurring multi-word phrases.

We’ll pick up this topic next week with more examples of great keyword-targeted content keyword lists.

Related reading

Stock photograph of a person in a suit with Facebook likes in bubbles floating past. The person is raising their index finger to touch one of the bubbles.

Simple Share Buttons