One of the best kept secrets of the Web is the fact that about 50% of search queries on any given day have never been entered into a search engine before. People type all kinds of crazy stuff into search engines. Often, the more specific and unique the keyword, the better it converts.

This means that the keyword data that shows up in Google Analytics should be a gold mine of opportunity to find new keywords as well as negative keywords. Google AdWords’ broad matching option includes an “expanded matching” feature that pulls in all kind of related keyword strings. Expanded matching captures the good, the bad, and the ugly keywords that you may or may not want in your campaign.

The problem is that Google Analytics does not give the data on the keywords that you need to build out your campaigns and target them better. It’s strange, but true. If you look in your paid keywords reports in Google Analytics for AdWords, you will only see the keywords in your campaigns that pulled the visitor to your site – you will not see the actual search query. For example, if your keyword in your Adwords campaign is “leather dog collar” on broad match, and this keyword triggered a search query for “suede puppy collar,” you will only see the keyword in Google Analytics show up as “leather dog collar.” You will have no clue what the real search query was, and whether you want to add that search query as a new keyword or exclude it as a negative keyword.

Google does offer a “Search Query Report” as part of its AdWords reporting options. Unfortunately, this report is filled with useless data such as “79 other search queries.” This report has become worse with the passage of time. It’s still useful, but who knows what’s missing from it?

The actual search query data is invaluable information for optimizing any AdWords campaign. Search query data is essential for true AdWords optimization for the following reasons:

1. It gives you new keyword ideas to expand upon and include in your campaign.
2. It gives you negative keyword ideas to filter out keyword queries that you do not want.
3. It offers the opportunity to increase your CTR and your quality scores when the two items above are effectively implemented.
4. It creates an automatic geo-targeting filter when combined with a state name in the keyword and a target radius.*

If you want to get this data, you best not look for how to get it in the Google Analytics help area, nor will you find it in the Google Analytics Blog. Instead, you will have to follow this link to a little known piece of java script that will pull that data for you. Many thanks to ROI Revolution for making this available to the public.

The only real question is why, with all the bells and whistles Google Analytics offers, why, with all the concern about quality at Google, why would Google not give us access to this critical data without using this complicated workaround? I can think of a few reasons — and they all end up making Google’s management look greedy. The first thing you will notice when you install the this work-around script is all the irrelevant keywords being triggered by your broad match keywords.

I can tell you first hand that having this data will not only save you about 10% on your ad spending, it will also generate about another 10% to 20% in sales when you start adding some great keywords. The best research I do for my clients is looking at what keywords my visitors use to get to a site and then look at what happens after they enter the site with that keyword. One good keyword idea can generate dozens of other keywords and landing page ideas and can result in a nice bump in sales.

Another good tool for this kind of research is hittail.com.

* When you combine a state name or abbreviation in conjunction with a keyword in a broad match, AdWords will automatically pull all city names in that state as an “expanded match.” This means that if you use these keywords with a target radius, you will pick up hundreds of small town and municipality keywords automatically. If you can see the actual queries, then you can put negatives in your campaign to filter out those names of towns and cities that creep into results because of geo-targeting anomalies.