Content Optimization: Keyword Analytics Tools

Content optimization remains an important focus of any SEO (define) campaign. Because you’re probably adding content to your site or blog constantly, there are always new opportunities to improve overall visibility on search engines. But to be found for the right words on any site, you must undergo regular, extensive keyword research.

You must understand the relationships between different words people use when they search for your goods and services, and how to bake these phrases into your site’s content. In my last column, “Content Optimization: Keyword Suggestion Tools,” I examined how to build out a keyword list and some common, typically free, keyword suggestion tools, such as Quintura, Google Suggest, Yahoo Search Suggest, Google Trends, and the Google AdWords Keyword Tool.

Keyword suggestion tools help you understand what words are used in search queries and what phrases are associated with those words. Keyword analytical tools provide some measure of keyword suggestion functionality, as well as the ability to understand the competition levels for specific search terms. More importantly, keyword analytics tools help you determine the relative size of the search referral market associated with specific keywords and phrases.

Yet many differences exist between keyword analysis tools and keyword suggestion tools. For example, Google tool data is compiled from the Google Search Network of sites, while subscription-based keyword analysis tools such as Keyword Discovery or Wordtracker use much broader databases of information that can be sliced, diced, and refined in many different ways.

In Wordtracker, keyword data is compiled from major metacrawlers, such as MetaCrawler and Dogpile. When performing keyword research in Wordtracker, the program “counts” the number of times a particular keyword or phrase has appeared in its database of 316,558,816 words. For example, a count of 150 means that a particular word has appeared 150 times in the Wordtracker database over a 100-day average.

Wordtracker contends that metacrawlers search data matches search queries of the major search engines very closely, and that the data isn’t skewed from software robots that continually check Web site rankings and pay-per-bid positions. This is an interesting point because Keyword Discovery uses a global premium database that contains over 4.4 billion searches based purely on user panel data to avoid the same types of data distortion.

Both tools strive to not inflate keyword search data in order to provide you with a more accurate estimate of available search referral traffic, even though the original sources of their data sets are very different. Both tools also provide a multitude of ways to slice and dice their data. Let’s put some numbers side-by-side to exemplify just how different some of these tools are.

To keep things simple, let’s look at the keyword forecasts for “dog food” from the Google AdWords Suggestion Tool, Keyword Discovery, and Wordtracker. (Note: All results are for what can be considered a broad match on U.S. search or user data.)

On average, Google predicts there will be some 1,395 searches for “dog food” each day. Keyword Discovery anticipates only 1,088 searches on a daily basis this time of the year, and Wordtracker suggests the potential search referral market for “dog food” is more than twice that number. Notice the connotation about Keyword Discovery’s daily prediction for this time of the year? Well, that’s where Keyword Discovery slathers on some great features because its algorithm as applied to its data set can account for seasonality.

When performing keyword research in Keyword Discovery you can review the seasonality of words and phrases used in search queries on a historical basis, monthly estimates or as part of an annual trend, as well as provide insight into search engine market share.

Toward this end, Keyword Discovery and Wordtracker data sets can be readily sliced and diced to estimate only Google search referral traffic, or any other major engine for that matter. In our example, the Wordtracker daily estimate for Google’s “dog food” search queries is 1,043, or 47 percent of the Daily Prediction data. In Keyword Discovery, however, Google has about 67 percent of its Average Daily estimate, so Keyword Discovery predicts that 738 “dog food” searches will be made in Google each day.

While this might not appear to be a vast difference in search referral traffic, consider how the data translates to an average 30-day monthly prediction. Google AdWords predicts between 37,200 to 46,500 clicks for top-positioned ads for “dog food”. Meanwhile, Keyword Discovery estimates there will be some 22,140 “dog food” searches this month, while Wordtracker projects about 31,290 “dog food” searches on average each month.

The accuracy of these estimates depends on your positioning, of course. You should always compare the actual data for top performing keywords against the predictions for a reality check.

Next time, I’ll examine what the keyword analytics tools are referring to for the Keyword Effectiveness Index (KEI) and occurrence data.

In the meantime, remember that generally speaking, the number of searches performed for each keyword will be much higher in Google than Keyword Discovery and usually Wordtracker. That’s because the Google AdWords Tool provides aggregated broadly matched totals by default, while Keyword Discovery and Wordtracker tend to be less motivated to sell you more ads that drive your bid pricing upward.

Either way, all of the data sets are relative. For starters, seek the most relevant, highly trafficked terms and phrases already reflected in your site’s content. As we continue through the content optimization process, you should know that you’re well on your way to achieving optimal results if you just put a little time into doing diligent keyword research.

Related reading

Brand Top Level Domains