Google Blogoscoped


WikiAnswers' Search Engine Optimization  (View post)

Zoran [PersonRank 3]

Monday, November 26, 2007
13 years ago8,138 views

One of my friends is running "pets website" for pet owners etc. and he have made "dynamic many" when new breed is submitted from some of the owners new link is added to footer many cloud. He did not think much about misspellings nor Google algorithms. With time that "cloud" become quite big reached over the size of the rest of page... and his site which was getting decent amount of organic traffic just dropped to 30-80 visits per day from very very looong tail keywords. He assumes that "cloud" could be reason... nothing suspicious was done there, no link selling, not MFA there was adSense but not very aggressive and users still use that site, only new members are joining in very low rate, and that is not good to community "always the same faces"...

Any taught or bad experiences with automatic menu creation or tag clouds?

Marcin Sochacki (Wanted) [PersonRank 10]

13 years ago #

[put at-character here]Philipp:

The WikiAnswers' footnote IMO is a plain and blatant keyword stuffing seo trick. They most likely analyze search engine referers and extract user queries to rank for misspellings of the original keywords.

Tadeusz Szewczyk [PersonRank 10]

13 years ago #

It's keyword stuffing, a well known method of search engine spam or misunderstood SEO.

Stephen Tordoff [PersonRank 10]

13 years ago #

It's not hidden, but it is in the page footer, so I'd say it is definately a SEO attempt

Philipp Lenssen [PersonRank 10]

13 years ago #

> s/requite/require/

Thanks, fixed.

Philipp Lenssen [PersonRank 10]

13 years ago #

[Edit: Grammar correction in the post: "Google's self-proclaimed litmus test for whether something is search engine spam is to check whether ..." – changed to --> "Google's self-proclaimed litmus test for whether something is search engine spam is to check if ..."]

J. McNair [PersonRank 10]

13 years ago #

Um, yeah, this is old fashioned keyword stuffing, obsoleted by linkfarms for 5 minutes, now back and better than ever.

At least they're somewhat honest about it and not using CSS or JS to hide it. This makes it about 1% less of a crappy waste of bandwidth

Still, don't most search engines know about misspellings now? I mean, MS Live Search has progressed to automatically correcting spellings in queries for common terms before actually querying anything. Arguably good content writers will bother to at least spell-check.

Crystal Williams [PersonRank 0]

13 years ago #

Hey Philipp!

My name is Crystal Williams, and I am the Community Coordinator for WikiAnswers. I wanted to take a moment to respond to your blog post regarding WikiAnswers Search Engine Optimization and Google, and also address some of your readers’ comments. As you pointed out, WikiAnswers is hugely popular and growing every single day. The stats speak for themselves. Being built on the wiki concept, however, some confusion comes into play. This is where I need to correct you on the so-called search engine spam or, rather, keyword stuffing. Not so. What you are seeing at the bottom of Q&A pages are actually similar questions (referred to as alternates), which aid our visitors in finding answers. Their placement was chosen as a convenient and out-of-sight location, as not to clutter the primary Q&As and related features – while also helping search engines know what the pages are all about. We are not in any way trying to cheat the system or Google, but rather trying to build a quality and permanent information resource.

Instead of having hundreds of questions all asking the same thing scattered about the site, we have the ability to merge together like Q&As. These are ‘manually’ created (they are man-made) by our users in one of several ways and are critical to the way in which WikiAnswers functions:

1) Contributors create them when they say a question is the same-as an existing one during the Ask process, and

2) Alternates can be added at anytime using the 'Edit Alternate Wordings' link located on the left in the menu area.

I should point out that the question that you referred to in your post, “Which breed of horse originated in Czechoslovakia?” was maliciously spammed and has since been corrected. Thank you for bringing this to our attention. We rely on a dedicated team of volunteers and watchful visitors, such as yourself, to clean up spam and stop vandalism in its tracks. It can get a bit overwhelming at times, but we are working hard to keep garbage off the site.

As a final word, I’m glad that you think we have a neat concept, and would love to see you give it a ‘hands-on’ try. We do have a relatively new Blogging category that might interest you. And I would be willing to mentor you, so that you learn how to use the site effectively and to the best of its ability.

Enjoy your day.


Crystal Williams
WikiAnswers Community Coordinator

Roger Browne [PersonRank 10]

13 years ago #

> Still, don't most search engines know about misspellings now?

Here are the top 10 queries that lead to (according to Google Webmaster Tools):

best bey
hampster ball
human hamster ball
buddy holly glasses
decor market

Notice that more than half of these are mis-spellings?

There was a time when I would carefully correct any mis-spellings that I noticed on the site. Now, if a user posts a mis-spelling, I often leave it. I figure if the users are spelling a term a certain way, then so are the people who are searching for that term.

Philipp Lenssen [PersonRank 10]

13 years ago #

Thanks for your information Crystal, I added an update to the article.

J. McNair [PersonRank 10]

13 years ago #

[put at-character here]Crystal Williams
Sorry for jumping on your site. It just seems like there should be a smarter way to solve at least PART of your problem (minor spelling and grammar mistakes). Even if it's as brain dead as running hunspell over all questions asked and then using the results, it could be an improvement over manually including "slightly off" questions. Your site's method is actually pretty clever for similar questions using different words/phrasing.

[put at-character here]Roger Browne
Point taken and correction appreciated. I guess I have just become a fan of how Microsoft's Live Search does it now. If it turns out I wasn't looking looking for the "corrected" query, one click takes me to my "original" query.

I think it shows an understanding of most search engine users. They don't know exactly what to type in to get what they want, and the query itself is never perfectly reliable because we're human. Thus all engines have some DWIMmery*.

It's a small, clever refinement of Google's own "Did you mean...?", which I use quite frequently. Mind, Google, Yahoo and Ask could probably implement this quickly, seeing as they all have spell check searching, but Live did it first and good on them.

*Do What I Meant (not what I said/typed)

Norm Drazin [PersonRank 0]

13 years ago #

If I'm understanding the WikiAnswers position right, they are doing something else that's interesting. Let's assume that the misspellings aren't necessary because the heavy lifting is done on, say, Google's side. That stuff is simply a waste of space and they ought to remove them, for their own good. But what Crystal Williams is saying is that these lists (the different ways of asking the same question, which may be use entirely different wordings) are based on what real people are doing on the site. It's like a log. They are recording it when someone points to a question and says "yeah, that's what I meant". That sounds like a creative way to use a community's wisdom. Assuming they are using this on the site itself when someone asks a question, this is merely exposing how they did it. Not quite "for humans" as in the original critique, but not really trickery either. Fascinating dilemma. If it helps visitors find bona fide matches based on previous visitors' tips, it sounds like it's on the level.

Stephen Tordoff [PersonRank 10]

13 years ago #

This footer still exists on the "horses" Q&A, but in a shorten form.

No sure if this is new, but there is also a tool tip explaining how the alternatives are generated

Sam Freedom [PersonRank 0]

13 years ago #

I'm not a big SEO geek or WikiAnswers user but just from Crystal's explanation, it sounds like someone could create 20-30 email accounts and then go around creating multiple similar entries for their chosen keywords. I don't know enough about it to know if they could point these back to external sites somehow (via the answer?) but if one could, then the system could be gamed.

Let's just go create 30-40 similar but different entries on "How to Game WikiAnswers" and have the answers point back here.

Yitz Forst [PersonRank 0]

13 years ago #

Sam, why not go try instead of posting theories?

You'll be amazed.

Philipp Lenssen [PersonRank 10]

13 years ago #

> then the system could be gamed.

To whose benefit? Creating longer footers benefits WikiAnswers, as it's more optimized (one might argue blackhat/ penalize-able) SEO.

Miguel C. [PersonRank 0]

13 years ago #

I enjoyed your post. AdAge just published a story related to SEO and Social Meda optimization. Matt Creamer, one of their editors, writes about his experience being "optimized" by SEO and SMM. You might enjoy the article:

[unlinked, just in case]

This thread is locked as it's old... but you can create a new thread in the forum. 

Forum home


Blog  |  Forum     more >> Archive | Feed | Google's blogs | About


This site unofficially covers Google™ and more with some rights reserved. Join our forum!