Google Suggest - The facts behind the results

The main feature associated with Google Suggest is a dynamically generated list that suggest keyword phrases and search strings in a drop down menu. The results of the list change as a user types in their keyword phrase, offering suggestions in accord with the topic. The purpose behind Google Suggest is to provide the user with alternate keyword phrases. The purpose is straight forward, but how the list is generated has spurred a lot of debate. Furthermore, there is a number called, “results”, displayed to the right of each suggestion. The meaning behind the “results” has generated many debates on what the term “results” means.

Arguably, to understand Google Suggest we have to know two things: what do the numbers associated with the “results” mean and how is Google generating the suggestions? First I will define what “results” mean:

the results are the number of competing web pages for that specific keyword phrase.

Next, how are the suggested phrases generated and how is the order determined?

Suggestions are determined by keyword popularity. The more popular a phrase, the higher on the list that phrase will be positioned.

This makes sense when a list is analyzed. If you open Google Suggest and start typing you will notice the “results” are not determining the suggested keyword positions. For example, if you start to type “website desi” (”desi is not a typo, I’m only entering a partial word), you will see “website designers” is listed second with 29,200,000 results. However, “website designs” is listed fifth with 49,900,000 results. So why are keyword phrases with lower “results” higher on the list?
This is because the list is being generated off of keyword popularity; which are not the “results”. To support the keyword theory, I needed to find more information about each suggested keyword. I used Wordtracker and SeoBook to help determine their relative popularity. What I found supports the theory that the keyword’s popularity is responsible for the position. Specifically, “website designers” had #67 searches over the last 90 days, compared to “website designs”; which had #31 searches over the last 90 days. Therefore I feel this supports why “website designers” is above “website designs”; regardless of “website designs” having a higher “results” number.

These results can be duplicated for any query, so give it a try. I want to make two remarks when you test my theory. First, using “website design”, as an example. There are dozens of phrases more popular than “website designer”. However, Google Search is looking for the next phrase that starts with the original query, “website design”. So phrases like “business website design” and “web design” are ignored. Second, when you start typing in multiple keyword phrases be careful when analyzing your keyword popularity. Once you compete a word, you must begin to type the next letter of the following word. Otherwise, “website design” will generate results that are not consistent with keyword popularity. However, typing, “website design c” will bring up a new set of suggestions that are arranged according to popularity, because I started the next word.

So now that I have demonstrated how keyword phrases are responsible for order, I want to define the number associated with “results”. First if we consider how Google has historically defined and used “results”, we can put some weight in knowing Google has never used the term “results” to describe searches. To demonstrate how Google uses results, do a search in Google and look to the top right of Google’s output. In a light blue bar is, “Results 1-10 of about (number of web sites).

If this is not enough proof for you, go to SEObook’s keyword tools. Type in a sample search and go to the fourth column from the right, it is called “Google Suggest”. Under the Google Suggest column, go to a row with some results, click the G Suggest link. This will bring up another page. Note the second column is, “Competing Pages”. Take note of the number associated with one of the keywords and go back to Google Suggest –the “Competing Pages” number will match the “results” number in Google Suggest.

Another test would be to type in your name, or some other (presumably) small search phrase. In my case Gary Reichel has #187,000 results. First, I don’t think that many people are searching for me, or anyone with my family name, but it is a reasonable assumption that there are #187,000 web pages with the names Gary Reichel. Especially when you search Google Suggest and note the description tag related to many of the results. You will see Gary is bold, but it is another person with the last name Reichel that is bold. Second, if keyword popularity was associated with results, there would be no reference as to what time frame these results were compiled. Therefore, the “results” would be useless.

Considering Google Suggest in these terms, you may begin to see how it can be used to do a general keyword search.

That is, associating the results as a number of web pages competing for a keyword phrase, can now be compared to lower competition web pages (results) that have greater keyword popularity (higher in the list)! Suddenly we have a nice tool for search engine optimization.

I hope this clears up how Google Suggest is basing suggestions. As far as using Google Suggest for search engine optimization, I think Google has developed a nice tool for webmasters and optimizers to use as another method for researching keyword phrases.

If you don’t agree I would enjoy a comment to the effect. If you do agree, well …how about two comments, and a high-five.

Bookmark It

Hide Sites