Mining Quora for Question Research
Traditional keyword research is outdated. In the past, a webmaster or SEO would use a tool like keyword planner to find a keyword related to their niche that had a lot of volume. That was enough to get started. Then the game changed and the tools became more sophisticated to keep up.
Tool providers like Moz started adding competition metrics to show how easy it is to rank for a specific keyword. In 2014, Google’s “Pigeon” algorithm made local much more important, so checking queries for implicit and explicit local intent became an important part of keyword research. Good SEOs began to understand that long tail searches were more likely to convert, so they began researching and pursuing those.
Each change to Google’s algorithm has made keyword research more complex. There are two algorithms specifically that impact how I do keyword research on a very basic level. Those two are Hummingbird and Rankbrain.
This post could get really long if I dive into how these two updates changed keyword research, but let’s keep it short and sweet and just say that searcher intent is now extremely important. If your page is not a good fit for the search, it does not matter how many backlinks you build to it.
It also so happens that traditional keyword research tools are not as adept at discovering what people want to know and what resonates with them. Keyword research tools are good at researching keywords, which are the fragmented chunks of sentences you plop into a search engine because you have to dumb it down for the machine. Those same tools aren’t so great at question research, finding out what questions people specifically want answers to.
That’s what Quora is for. Any Q&A site can work, so while Quora is popular and widely used, we’ll focus on that. If Yahoo answers ever makes a comeback, I’ll mine that in the same way I do Quora. For this article, let’s use a limousine service as a case study.
The first thing we’re going to do is find questions that people are asking using advanced search operators. First we’re going to start with something specific then get broader.
This tells Google that I only want to see results from Quora that have limo in the title. I immediately get a few results that look interesting.
I’ll want to take my time exploring these, but it bears mentioning that you can adapt your search operator to expand your research if you’re not getting enough useful information on the first go around. A few ideas are to try varying your head term, like “limousine” instead of “limo”. You can also change it up to get broader results using intext, instead of intitle.
Type into Google search: site:quora.com intext:”limo”
This tells Google you want to see any search result that includes “limo” in the content of the page. You will get broader topic ideas doing this, but with the reduced specificity you will also get some junk.
Measure Views Instead of Volume
Normal keyword research tools measure historical search volume, and provide it as an estimate. It’s good info to have, but as an estimate it doesn’t always accurately show how popular a particular question or answer may be. Quora, on the other hand, shows exact views that an answer gets, which gives you an idea of the popularity of both the question and the answer.
This can be just as or more useful than search volume. Most keyword tools operate on one of two systems when calculating volume. Many use data from keyword planner, which creates estimates using broad match. This means if Google’s algorithm thinks the keywords are related (it gets this wrong a lot), it bundles their search volumes together. Then there are tools that use clickstream data, which is used to show volume estimates based on exact match. They say it’s 95% accurate, but compared to what? Compared to keyword planner data, of course! That doesn’t mean the insights aren’t useful, it just means we have to always be cautious when we’re making our estimates and predictions, and find other methods that drive insights as well.
We can compensate by checking out Quora to see how interested people may actually be in a question. This is also subjective, because views can be influenced by things like questions being featured, or who responds to the question, but it’s still valid to say that if no one is interested in the answer, they likely won’t view it. So let’s take a look at the first result from above, “How come you’re allowed to drink in a limo” and see what Google’s Keyword Planner has to say about it.
Okay, apparently no one wants to know this. What about Moz, since they use clickstream data?
Okay, both estimates show no one is asking this question. Now for the million dollar question, how many people are viewing this on Quora?
Okay, 532 views since June 21. This is one of the lower view counts, but it’s also very recent. You can do the math here, that’s roughly 200 or more views per month, which indicates a demand for the answer. I like to then compare this to the Question Stats box in the lower right.
This question has 8,766 views since Feb 2012. We could break it down with some simple math to about 110 views/month, but that’s not totally accurate. Who answers the question impacts the view count. But the important thing is that we’re identifying from these two pieces of information that people are interested and continue to be interested in the answer to this question.
Editors Note: While we’re being a little flippant, it’s entirely possible no one is actually searching this on Google, which is why there is no search volume associated with it. That doesn’t mean that answering the question won’t help you rank. Creating useful, engaging content that answers questions no one else is adequately answering is not only link worthy, but sends positive user signals to Google.Brett Elliott
Traditional keyword research didn’t pick this up. This could be for many reasons, like the tendency for searchers to type fragmented sentences into Google. It’s for insights like this that we want to mine Q&A sites like Quora. And while it may not be searched directly in Google, it’s probably an extremely relevant bit of information for anyone searching for a party limo or renting a limo for a wedding. That added value is rewarded by search engines as they evaluate how engaging your content is.
As SEOs we’ve used the related searches and people also ask features on Google for years to do keyword research. Quora has a section itself that might help you find more content ideas. Some of these are junk, but others are more insightful. The question about drinking in a limo will yield a lot of related questions about drinking. The related questions for “what is so special about traveling in a limo” is much more useful for us.
You can really go down a rabbit hole with this, so be careful. Set your hourglass upside down or something.
Bonus: It’s Easier to Write the Content
If you’re a content marketer you know what a chore it can be to research the answers to a pile of questions when you’re writing some long form copy. Luckily, you have a set of answers built into Quora, many of which will cite a source you can examine.
You’ll also run across a lot of bad answers that perpetuate myths, or answers that don’t answer the question at all. Reading bad responses can actually help you write better content. I’ve highlighted incorrect information before to educate my readers so they know what to look out for, which helps establish your authority and provides something useful in return.
This is about Question Research
This method of keyword research is more about researching questions than it is about keyword research. It can’t replace what we do with our keyword research tools, but it does supplement it nicely. Instead of focusing on TF-IDF, LSI, or any other buzzword SEO trick of the day, I look for things I know people are asking related to the topic. By answering these questions fully and thoughtfully I’ve had a lot of success hitting page one.
As always, if you’re interested in SEO services, please contact us using the form below. Thanks for reading!