Wordtracker Keyword Questions
Wordtracker has released an interesting new tool recently called "Keyword Questions". It's a free tool that lets SEM's find specific questions that people type into search engines.
After typing a few keywords in the search box the Keyword Questions Tool will search the Wordtracker database for search phrases consisting of full sentences that contain your keywords. We get to see up to 100 results and each result is shown with the number of times the question was asked in the last 140 days.
You can use the results to create landing pages centered around a particular question or create a FAQ page which addresses the questions provided by the tool.
According to Ken McGaffin, CMO of Wordtracker.com: "People have a ton of questions about all sorts of things. They want answers and many of them will go to a search engine to find them. By providing the answers to their questions, you’ll create interesting copy and pick up a lot of relevant traffic for your website." (source)
I like the idea, but I'm a bit worried about the size of their database. Wordtracker.com uses search statistics from the meta search engines MetaCrawler and Dogpile to fuel their tools. People who surf directly to Google or Yahoo will not be counted. In the Netherlands Google accounts for almost 95% of all searches, so we would be looking at a tiny fraction of the Dutch search volume.
I also heard at SES San Jose that only 5% of all searches are based on full sentences. Search engines often have trouble processing English sentences and will present irrelevant results. When a SES attendee asked Erik Collier (of Ask.com) when search engines would be able to reliably handle these kinds of searches, he said: "It's gonna take years".
So will the figure of 5% go up or down in the next couple of years? If it goes down the Keyword Questions results will become less relevant over the coming years. Also keep in mind that the people doing full phrase searches might not be your target audience. If you're trying to sell to the 18-24 demographic you will probably not find this tool very helpful.
But anyway, time to put the tool through a test. In my test scenario I'm trying to optimize a gift store. A search for "christmas shopping" gives these results:
Only 6 searches? That's not very helpful. Compare that to these Google trends:
When I run "christmas shopping" through Google Keywords instead I get 200 keyword suggestions. Admittedly these are not full sentences, but I still get useful long-tail suggestions like:
[location] christmas shopping opening times
The reason that Keyword Questions does not find many results is because of how the tool was designed. It combines your search keywords with the words what, why, when, how, where and who, and then searches its database for questions matching these keywords. Questions that do not contain at least one of these words will not be found.
Most of my Keyword Questions searches for the above word combinations return 0 results but when I try the single word "gifts" I finally get some useful suggestions:
Not bad. So can we build landing pages around the top 2 questions? Before we do, let's check out the October search volume reported by Google for combinations of "gifts for...":
"Gifts for him" and "gifts for men" score the same, but "gifts for her" scores almost three times as good as "gifts for women". And "gifts for the man" does not appear anywhere. Google also tells us (not visible in the image) that the plural "gifts" scores a lot better than "gift". So when writing SEO copy a good headline would be "Gifts for her - the woman who has everything", instead of the top suggestion offered by Keyword Questions. And did you spot the spelling mistake?
My point is that you cannot blindly trust the Keyword Questions tool to provide good SEO copy. You always need to cross reference the wording of the questions with the results from Google to make sure that your copy contains many high volume keywords.
Searching for short terms like "woman" or "her" result in a surprisingly high number of porn-related questions. I also see many spelling mistakes ("scare" instead of "scar", "expercise" instead of "exercise"). Many of these spelling mistakes rank a lot better than their corrected versions. This all makes me wonder how reliable the underlying database is.
But Keyword Questions does come up with unusual suggestions that might help you identify niches. Further down the list we find:
Conclusion: it's an original idea but this tool is not very useful at the moment. The relatively small database, the small percentage of searches involving full sentences and the short list of pronouns really limit the number of useful results that you will get out of this tool.
However, Keyword Questions is good as a brainstorming tool and I have to say, it's also excellent link bait.
Tagcloudstreetview microsoft analytics smx search engine interview sea google maps google earth realtime ses youtube london linkbuilding bing advertising business mobile event social media gmail search facebook twitter video matt cutts marketing yandex adwords china sem search engine strategies searchcowboys apple images funny viral app maps blog