• Dicebat

    Thanks for sharing your insight, Shari. You also made a good point. I very often see sites that are optimized for search engines instead of the user. Fortunately enough, that’s the moment I get to fix it :)

  • http://twitter.com/sharithurow sharithurow

    Hello Dicebat.  I don’t know what your first experience was with user-centered design and usability testing. Mine was a very, very humbling experience.

    It was an ecommerce site. I was dumbfounded when I observed users/searchers not do what I did…and I fit the client’s target audience (primary persona). So my personal mental model was inaccurate.

    I was also surprised when I compared the keyword research data with the final usability data. They didn’t quite align in the way I thought they would. 

    Very eye opening experience. I’m glad I had it, though. 

    I don’t expect SEOs to become usability professionals. But I do wish SEOs would work with usability professionals so that they can see if or when usability data and keyword research data possibly conflict.

    Thanks for the comment!

  • http://twitter.com/seo33 SEO 33

    Always insightful , Shari. I think people working with PPC advertising are much better understanding the logic in you article because understanding user intent is a must and because  
    they get instant feedback of changes they make and have the ability to test. Many SEO I meet are more concerned about keyword search volume and ranking. My advice to every SEO is sit with the PPC guys to understand the target audience.

  • http://twitter.com/sharithurow sharithurow

    Hi SEO 33. Thank you for your comments.

    Hmmm, I wouldn’t conclude that PPC guys have more usability insight. Sure, they get data, but they do not necessarily get the data in the proper context. 

    That’s what happened with that last example. I had all of this data from web analytics, PPC ads, and keyword research tools. None of this data supported the fact that this industry’s target audience organized and labeled content by target audience.

    That was Donna Spencer’s point as well. In her webinar, she stated that she wouldn’t have the “why” without the face-to-face interactions.

    I would never architect a website based on PPC data alone. Context is so important. I agree with you — search volume and ranking seem to be more important to SEO professionals than other things. I think the result is designing, writing, coding, etc. for search engines, not searchers.

    Thank you again!

  • http://www.zagoumenov.com/ alexander zagoumenov

    Shari, thanks a lot for an insightful article! I do occasional keyword research projects and I agree that context and classification are very important in connection with volume and competition data from analytics.

    I’ve also done a few site focus groups and although these activities are very helpful, it usually costs the client a substantial sum of money. Although I’m not a fan of shortcuts, I have to resort to research shortcuts every once in a while to minimize client’s costs.

    One of those shortcuts is asking people at reception to fill out a keyword list that they would use to find my client’s business. This helps me to see what terms the actual target audience will use to find the site. Then, by applying the query classification I can break the lists further into valuable information for my keyword research.

    My question is: what other qualitative keyword research shortcuts are there, based on your experience? Again, I’m not a fan of shortcuts, but it’s great to have a few of these in your arsenal as an SEO, especially dealing with SMBs. Thanks in advance for your reply!

  • http://twitter.com/MaryKayLofurno Mary Kay Lofurno

    Shari – Great article as always.  Agree with the query classification and searcher personas.. I have a different question as you know, Google just introduced its knowledge graph.  My dev director and I are starting to investigate the potential for coding into the family of sites we have where appropriate.  A technology we are looking at, opencalasis has an extension for one of our CMS’s [we use a few different ones]  I am told this technology has a ‘machine’ type card sort mechanism.

    So, given your comments above, especially about card sorting, its seems like if ‘query classification and searcher personas’ were not originally considered that the data fed and coded in to a knowledge graph [semantic web way] would be even more…flat..not sure if its the best word…

    Thanks, Mary Kay

  • http://twitter.com/sharithurow sharithurow

    Hi guys-

    Alexander, I understand shortcuts, because not every SEO is going to run out and become a usability professional like I did (and it took years of classes and experience with mentors). But the one thing I do know about testing is that you ABSOLUTELY have to test people who fit the persona or profile. 

    If you don’t, you are probably not getting the right mental model. And another thing that is really hard? When you ask others to describe your client’s services? You might be leading them with your questions and your responses. Also, are you giving them the exact same scenario? We have to do that for the data to be accurate.

    Practice practice practice.


    Mary Kay, Google is a technology company. They treat their technology data as the Holy Grail, IMHO. When someone from Google touts search-engine friendly design? He/she is touting good for Google, not necessarily good for users (though Googlers preach that). I have conducted too many usability tests over the past 10 years that completely contradict Google’s data, like the one I mentioned in the last part of the article.

    So I would accept their data for what it is, but conduct my own tests to validate if what they claim is true for your websites.

    I do not want to give the impression that search engine reps are always wrong. They are not. I have validated navigational, informational, transactional intent many times over. 

    I’m wary of anyone who says to do something for the user experience when he/she doesn’t define it.

    Great questions!

  • http://twitter.com/MaryKayLofurno Mary Kay Lofurno

    Thanks Shari – Yes, we are talking about testing the knowledge graph and semantic stuff with one of our content sites now as opencalasis has an extension for the CMS its on.  We also have a pilot on rel author running and planning them with the event & product rich snippets.  So we are doing that.  If you are at SMX NYC, lets try and catch up.  I hope all is well. Have a great summer. Mary Kay

  • http://twitter.com/sharithurow sharithurow

    Hi Mary Kay-

    As an FYI, I am getting mixed results with the rel=author and the image display with author photos. I think the individual results depends on the industry. 

    Hope to see you in NYC, too!