Updated Google Quality Ratings Guidelines (for Q&A sites)

Question

One of the hot topics of the last week in SEO was news that the confidential Google Quality Ratings Guidelines were re-written a short while back.

For those that don’t know, Google uses thousands of individuals (humans) to rate sites for quality, and the new guide helps those people rating sites with what makes a quality site.

One of the examples given was for;

http://wiki.answers.com/Q/Where_water_is_plentiful_soil_is_often_what&isLookUp=1#Q=where%20water%20is%20plentiful%20soil%20is%20often%20what

"This is a page from a Q&A site with an unanswered question. Q&A pages exist to answer user questions. Pages with unanswered questions should generally be rated Low because they have little MC (just a question and no answer) and don’t achieve their purpose well. Misleading or potentially deceptive design makes it hard to tell that there’s no answer, making this page a poor user experience. In addition to having no answer, this page has Ads and links to other questions (misleadingly labeled as “Relevant answers”) displayed prominently, which users may mistake for answers to the question. It takes a moment to notice that this page actually has no answer"

I suggest that for the purposes of SEO, questions with no answers should receive a meta robots noindex directive, until an answer is given.

No answer;

<meta name="robots" content="noindex, follow" />

</head>

With answer;

(no meta robots directive (which means index,follow).

</head>

 

*I've implemented a noindex on tag pages now* this was done using;

---
    header('X-Robots-Tag: noindex,follow');

    if (!defined('QA_VERSION')) { // don't allow this page to be requested directly from browser
        header('Location: ../');
        exit;
    }

 

within the qa-page-tag.php file.

Does anyone know how to add this to pages with a question and no answer, but, remove it when the question does get an answer?

Comments

My thought to "Google uses thousands of individuals (humans) to rate sites for quality" -- Really? They should have to read the content and then evaluate it. There are billions over billions of websites... where did you read this?

---

It's been known for years now, (2007 NYTimes link below).

http://bits.blogs.nytimes.com/2007/12/18/the-people-inside-googles-black-box/?_php=true&_type=blogs&_r=0

"Marissa Mayer, in a talk in June, said the company had more than 10,000 people checking its search quality"

There are machine learning algorithms to deal with thin and low quality content, ('Panda'). However, these kinds of changes are checked via human reviewers. What the leaked reviewer guidelines give us is an insight into the kinds of content Google is looking to deliver to it's users and it's fairly clear questions without answers is not one of them. Eventually, Google will stop asking humans to help identify these kinds of sites as they seek to algorithmically demote this kind of content.

via http://www.thesempost.com/google-rewrites-quality-rating-guide-seos-need-know/

"Hidden text

Google previously asked raters to investigate any white spaces at the right or bottom of the webpage if it seemed to be there for no reason as well as using Ctrl-A, turning off CSS and turning off javascript to identify hidden text.   But they are no longer asking raters to check for this anymore… again possibly because Google feels they can usually catch this algorithmically now."

Citation: http://searchengineland.com/googles-quality-rating-guide-leaked-version-five-196619

http://dejanseo.com.au/research/quality-raters-links/

---

oh yeah, I know people making a decent (not a lot , but better than being unemployed) doing QC for Google. Also, if you check Amazon's mech turks, you'll find a lot of the posting consist in looking at webpages...

---

Thanks for the info :)

Answer 1

This is very well thought and I'd have to say it is perfect for users who are looking for answers. As you said, it makes no sense for them to find a question with no answer.

However, there are 2 sides of that coin. In order for someone to provide an answer, a question should be searchable. Suppose you are looking for an answer on a question and you end up just finding the question with no answer. After a while you research and get to your own answer. Then you go back to the question and provide the answer you've found as a means of helping the community.

In the latter situation it wouldn't be possible to have find the question and hence, you couldn't have proposed the answer to it and user couldn't have benefit from that (or even improve that). I even think playing with dates and provide a mixed solution wouldn't be a good idea. EG: Suppose you say if a question is over a year old then you add the meta tag. This is still not good because old questions can still be absolutely valid and answers can be found to them.

I know this because I'm taking it from a real situation I've faced in StackOverflow (which I will use as a reference of QA sites). Some years ago I found this question (no answer had been given to it) http://stackoverflow.com/questions/1156759 . I was desperate to find a solution and I just didn't find it. After playing for a while I found the solution and posted it. It was a year after the question had been asked and even the user asking the question had been removed. Anyway, since then many people benefited from it and it wouldn't have been the case if I couldn't have found the question.

So, again, I like the idea and it makes sense. I'm not the one to evaluate whether this is more SEO friendly or not as I'm not an expert in that matter but I'm pretty sure this will, in a limited way, decrease user collaboration.

---

Answering your post edit question: Does anyone know how to add this to pages with a question and no answer, but, remove it when the question does get an answer?

You're basically hacking the core fixing this that way. It'd be better to just go to your qa-theme/<your-theme>/qa-theme.php file and then make the doctype function look this way:

function doctype() {

    qa_html_theme_base::doctype();

    if ($this->template === 'question' &&

        isset($this->content['a_list']['as']) &&

        count($this->content['a_list']['as']) == 0) {

        header('X-Robots-Tag: noindex,follow');

    }

}

 

It is still a bit quick and dirty but should work well enough.

Comments

Unfortunately Google don't agree with this logic and if SEO is an important part of your traffic mix, I suggest that those unanswered questions get a noindex. Which... doesn't stop the unanswered question being found onsite, just from being indexed and ranked by the search engines, (those that adhere to the meta robots directives).

It might decrease user collaboration, though you'll find already that Google is pretty smart about questions without answers algorithmically, (those pages aren't ranking very well anyway).

---

really good idea my friend :) !

---

@Simon I've just edited my answer to also tackle your edit

---

Thanks @pupi1985. I tried adding it as below and questions with answers (as opposed to what was expected, questions without answers) were sending noindex,follow in the header response.

            $this->output(
                '<SCRIPT TYPE="text/javascript"><!--',
                "if (typeof qa_wysiwyg_editor_config == 'object')",
                "\tqa_wysiwyg_editor_config.skin='kama';",
                '//--></SCRIPT>'
            );
        }
       
        function doctype()
        {
       
            qa_html_theme_base::doctype();
            if ($this->template === 'question' &&
            isset($this->content['a_list']['as']) &&
            count($this->content['a_list']['as']) > 0) {
            header('X-Robots-Tag: noindex,follow');
    }
}
       
        function head_css()
        {
            qa_ht


Any ideas?

All the best,

Simon.

---

My bad. It was not "> 0" but rather "== 0" (check the updated code)

---

Brilliant! Thanks!

---

i try add this code but no show you can help me for more info