Segmenting question queries for fun and profit.

In dealing with larger websites, there’s often little snippets of useful data hidden away in your analytics software.  When analysing the keywords which send traffic to a site,  it makes sense to segment that data to help make it easier to process, and understand. On a recent project of my own, I found that technique particularly useful when looking for question queries , and making sure that the website was providing the information necessary to answer the question, and either convert – or prevent a bounce.

What are question queries?

Simply put, question queries are the direct questions that your visitors are asking the search engines, and landing directly on your site.  e.g. “What colour is the sky?” – Google have been trying to improve this section of their algorithm,  (see this research paper from Googlers who realised that question queries, as opposed to keyword queries failed more often) – with a social question engine Vark (acquired) most likely being used to improve the results when question related queries are presented.  Interestingly, they also found that question queries are also formulated more often when other search terms failed, so this often presents webmasters with an opportunity to “plug the hole” so to speak, and provide searchers with the information they need.

Finding your own question queries?

The first thing to do, is to work out what constitutes a question. My own technique for this with Google Analytics, is to pull out all question words. For example “Who”, “What”, “Where”, “Why”, “When”,”Should”,”Can” are all words that constitutes a question when used in a long tail phrase – particularly when used at the start of the string. You can create a segment to show you these quite easily.

In Google Analytics Under “Manage Segments” > Go to, Create New Segment, then make your screen look like this:

Or alternatively,  I’ve shared the segment directly here. You should hopefully be able to click on that link and apply it to your own Google account. Once you have that in place, you can apply the filter, and see what terms come back. Plugging these back into Google will let you see how you are performing, how well your page answers the query, and where you can possibly improve.

If nothing else, this data alone could make up the basis of a frequently asked question page that takes visitors to other pages within your site that better answers the question.


Submit a Comment
  1. I love your perspective and approach. There are so many ways to analyze data and one thing I preach to all my clients is that you must allow your data to help you make your decisions and modifications of strategy. This is a nice detailed example of what can be used and more importantly should be used.

    I believe that it is this kind of expertise that can help businesses better leverage the power of inbound marketing. Well done!

    Daniel Shlifer

  2. This is a great segment, and the share link worked brilliantly.

    This and your social media segment are really good articles, would love to see more as time goes on.

    Many thanks

  3. This was great, thanks for taking the time to post with the screen shots. Just ran through it myself and there is some great info in the report that I would of looked past all too easily.

    I didnt notice the share link till the end but it was easy enough to run through manually. If I am able to leverage any info for a new landing page or FAQ page ill make sure I blog about the results and link back here.

Leave a Reply

Your email address will not be published. Required fields are marked *