Quality According to Google - Official "Guidance" on Panda Update

[netmeg]

Finally a post from Webmaster Central about what Google thinks about quality:

[googlewebmastercentral.blogspot.com...]

[tristanperry]

Yep, a very interesting read :) It's interesting to consider just how much Panda does use these pointers in their ranking algorithm.

I mean, "Would you be comfortable giving your credit card information to this site?" is probably something that'd be tricky for Google to accurately implement via an algorithm.

Whilst some of them ("Does the site have duplicate, overlapping, or redundant articles on the same or similar topics with slightly different keyword variations?") should be relatively easy for them.

However:

"Does this article have spelling, stylistic, or **factual errors**?"

"Does this article provide a complete or comprehensive description of the topic?"

"Does this article contain insightful analysis or interesting information that is beyond obvious?"

- if they are actually ranking signals, of course - would suggest that the Panda is fairly sophisticated and has had decently big leaps with respect to latent semantic analysis.

[falsepositive]

The international scrapers that rank above me would be sites I'd be deathly afraid to give my credit card to!

[londrum]

a lot of those things make sense when you're marking an english exam, but not when you're ranking pages on the web.

things like "Does the article provide original content or information, original reporting, original research, or original analysis?" is all well and good, but how does "original content" equate to quality?
just because something has never been said before doesn't mean that it's any good. it's practically impossible to rate the quality of a page by looking at whether the words are unique.
In fact... if you have 10,000 pages on one subject, and 9999 of them more or less say the same thing, and the other one has completely different words, then i would suggest that it's actually much more likely that its information is WRONG, rather than better.

"Does the article describe both sides of a story?"
Likewise with this. What has this got to do with quality? Most commentators are biased. Most newspapers too. Does that mean they are not worth reading?

"Is the content mass-produced by or outsourced to a large number of creators, or spread across a large network of sites...?"
This is totally wrong. Syndication is usually an indication of GOOD quality, not poor quality -- otherwise people wouldn't bother buying it.

[koan]

Finally some authoritative document from Google, there's a lot of good stuff in there. I don't think we can expect more than this from Google without them revealing too much.

[edited by: koan at 8:44 pm (utc) on May 6, 2011]

[whatson]

"One other specific piece of guidance we've offered is that low-quality content on some parts of a website can impact the whole site’s rankings"

So every single page must be perfect, or the whole site will suffer.

[Leosghost]

@netmeg ..Thank you for the link :)..just noticed the thread.

[jinxed]

Thanks - this was good to read.

[supercyberbob]

So much damage control. So little time.

[lucy24]

"Does this article have spelling, stylistic, or **factual errors**?"

"Does this article provide a complete or comprehensive description of the topic?"

"Does this article contain insightful analysis or interesting information that is beyond obvious?"

Google's scraping wikipedia?

[ken_b]

What the heck are these people smoking?

I believe I'll stick with their old advice and continue to build my site(s) and pages to meet the needs of my readers/visitors and not to make a search engine happy.

If that costs me some, most, or even all of my Google traffic, I can live with that.

[whatson]

I understand Ken, but this is my living and although I enjoy it, I still need to get paid.

[walkman]

Sorry but this works when a professor ranks a term paper, even then in SATs you get 3 raters.

No wonder Panda has had so many false positives.

[scooterdude]

Interesting article, it does put forward some of the type of content a decent search engine would want to rank,


However, to me, there is a disconnect between the target and the outcome i see in the SERPS

Plus I am pretty certain that the SERPS prove that they don't have an algo currently capable of delivering the stated objective,

Nevertheless, its a credible description of what might give webusers a good web experience.

[Shatner]

These guidelines seem as though they were written to address only competitors of the NY Times.

They aren't really applicable to all kinds of sites. The internet is more than newspaper articles.

I find that aspect of this a bit confusing.

But even applied to content sites which do fit into this mold they're building, I've seen many cases personally of sites that fill all these criteria and were still Pandalized. Not all, but many. The scraper outranking problem is the most obvious example of this.

So clearly there is MUCH more to it than this. Google isn't really being 100% honest. But at least it's something and I'm sure all of this plays at least a PART in what's going on even if it's not the full picture.

[walkman]

I re-read it carefully and on my hit site the only one that directly applies to me is :
"Does the site have duplicate, overlapping, or redundant articles on the same or similar topics with slightly different keyword variations?" and that's because of tags (removed them two months ago). Now "will you share with friends..is it better than... etc" depends on the user I guess. It's not the top site but it's not that bad either :)

But then my hit site is not an article site so, as Shatner said, there has to be more to it. A lot more.

This appears directly against content farmers but we know that eHow embarrassed Google by surviving it and others top content sites got the axe.

My last thought is that brands and big sites will be getting an even more major boost and others better have Plan B, C and D. Even if you get it right, by the time Google figures out their brand new algo and credits your site for any changes, it may be too late for your employees or even you. We are 2+ months in this already and ranking for most hit sites have only gone down again, even after making changes.

And no, Google hasn't gotten anywhere near right.

[Whitey]

Reading between the lines, Google is dividing it's emphasis up between informational or publishing sites, and transactional/business sites.

The guidelines provided sway strongly towards the informational / publishing . There is a "gray" overlap with business & transactional, and Google has it's sites firmly set on advertising and subscription revenues that conflict with it's existing business model.

Authorative niche articles , branded = longer life
E-commerce , comparison , business listings = shorter life

[ErnestHemingway]

My sites pass all the questions and much much more but still got slammed. It is crazy how you guys are operating Amit. No doubt Bing is catching up.

[Reno]

As one of the people who has criticized Google relentlessly for not defining "quality", I commend Amit Singhal for this blog entry today. Perhaps they are paying some attention to the webmaster community afterall.

As noted in the previous postings here, it may not make complete sense to us in every case, but at least it's a thoughtful response. My only critical comment is that the guidelines should have been posted 6 months prior to the Panda rollout.

Thanks netmeg for the link...

......................

[Pjman]

Perhaps they are paying some attention to the webmaster community afterall.

They better start smarting up and being more out and in the open about this. I'm not saying give us the keys to the castle, but at least give us an indication why Scrapers have #1,2,3,4 and I'm on page 2.

I a large number of people's eyes, they went from "Do No Evil!" to "If We Are Perceived As Doing Evil, Just Don't Talk About It And It Will Go Away!"

Even though they might not be sending as many visitors our way, many web masters have pretty strong followings from many channels. In a typical day 50,000 people without any of Google's help (referrals or Adwords), will read my content. Maybe 20% take some sort of action. If I start saying, I no longer trust G and their policies. Which I am on the verge of, it will surely turn off some people to G.

They need to continue to communicate well. They want happy users and happy publishers otherwise their market sahre will drop.

I think I did 10 total Bing searches prior to this year, mostly by accident. Now I do about 50-50. My readers know this.

[supercyberbob]

Speaking of not communicating.

I see no mention of a timeline between making changes and recovery of rankings, or why some webmasters have made changes and seen no results.

Also,

"Are the articles short, unsubstantial, or otherwise lacking in helpful specifics?"

That sounds like a description of Yahoo Answers, which seems to be untouched by Panda?

[Swanson]

This is press and relates nothing to do with what the algo is actually doing.

I have run 10 parallel tests on my existing and new sites and this information is invalid.

Larger sites clearly have a bigger problem with this change - it looks like some sort of percentage based calculation is at work and as a result larger sites tend to have bigger problems.

I don't want to give out the details of my tests as they actually gave me a clear insight into how to proceed ongoing however i will say this:

Pagerank is no longer a good indication of the ability of the site to rank well with even good unique content. The "quality" of the backlinks that comprise the pagerank is more of an issue. However, lower pagerank sites will struggle more than ever.

This update has clearly changed the "power" that links from certain types of sites hold - and I think this more than anything is the reason so many different types of sites are being "pandalised".

Yes, there may be onsite issues as discussed in other threads, however the real issue is that links from "lower quality" sites (that have been scored lower in the algo) seem to pass less juice than they did.

The result is that SEO based links - i.e. articles, blog posts, social bookmarks, paid links from thin sites - all now pass less juice (yet still report the same PR).

Brands rely less on these type of links and so are less affected.

What I think is now happening is that there is a scoring within pagerank - i.e. a hidden scoring that means the power of a XYZ pagerank link becomes clouded as it is based on the quality of (a) the source link and (b) the target page.

Just a theory - but it is as if there is now a QR (quality rank) scoring that is hidden from us but now is added to the pagerank algo.

[Swanson]

The net effect is a massive downgrade in the power of the existing links to sites.

It doesn't take much to destabalise the whole link profile that sites have - and if you are an SEO then you may be harder hit if you don't have a diverse link profile, but more worrying is that if you are an ecommerce business your link profile will probably be weak in any case as it is so hard to get stable high quality authority links.

You basically need to know SEO and link building more than ever, and you need more than one site.

If you have a solid link profile you can get away with lots more than the others - clearly ads, content etc is an issue but the more stable your links are the more you are ok.

[Whitey]

The net effect is a massive downgrade in the power of the existing links to sites.

Spot on & I agree that this makes large sites that pump out lot's of pages vulnerable. Amit Singhal is specific in this paragraph :

“One other specific piece of guidance we've offered is that low-quality content on some parts of a website can impact the whole site’s rankings, and thus removing low quality pages, merging or improving the content of individual shallow pages into more useful pages, or moving low quality pages to a different domain could eventually help the rankings of your higher-quality content.”

All this talk of "engagement" is important [ see Eric Enge's 20% theory ] , but not as important as ramping up the the basic fundaments. Good links , good neighbourhoods , quality content. Nothings changed fundamentally and sites that have fallen below the new threshold will have to spend a lot more time and effort to win back the trust they thought they had.

I'm thinking an automated time penalty has been added into the algo, otherwise we'd be seeing lot's of reports of quick improvements.

[edited by: Whitey at 2:07 am (utc) on May 7, 2011]

[walkman]

Just a theory - but it is as if there is now a QR (quality rank) scoring that is hidden from us but now is added to the pagerank algo.

Yep, I think we all agree on that. The question is how often is this score updated, how fairly can google calculate it, and what % of power does it have ?

[dazzlindonna]

shatner said:

These guidelines seem as though they were written to address only competitors of the NY Times.

They aren't really applicable to all kinds of sites. The internet is more than newspaper articles.

And there's the problem I see. My site that was hit hard really has very little to do with articles, and certainly isn't a site that anyone would need to worry about using a credit card on. Imagine a place that allows people to freely download original works of art, for instance. While the site has some text on it, it isn't meant to present factual information of anything, or present even one side of anything, much less two. It doesn't ask anyone to spend any money on the site, so there's no need to trust it with a credit card. There's actually very little in that post that has anything to do with my site that was hit, and yet it was. Whatever, Google.

[Reno]

The internet is more than newspaper articles.

In his bulleted points, Amit uses the word "article" at least 11 times, so it appears the advice is slanted towards information sites rather than product based ecom (though other points do address that).

how fairly can google calculate it

This is a crucial observation. There are information based sites and product based sites, and each would require different standards (it seems to me). Can an algorithm differentiate between the two, in regards to "quality"? Maybe this explains why I (and others) have taken a nosedive. If the algo is set to favor "articles", then those of us working in sales may continue taking a hit, at least until they come to the obvious conclusion that product based queries cannot be ranked in the same way as information requests. In those circumstances, putting an information page in front of a product page is almost certainly the wrong response, and thus only frustrates the searcher, no matter how much "article quality" the information page contains. Sounds simple, but can an algorithm figure that out?

..................

[ascensions]

I think I'm stupider for reading it. Thanks again Google.

[Whitey]

I do urge not to loose focus of Swanson's comment about Amit's declaration - sorry to repeat myself :

The net effect is a massive downgrade in the power of the existing links to sites.

There is nothing new about the principles of the new algorithmn - it is just the effectiveness of it.

Effected webmaster's are panicked and quick fixes will not work straight away. The factors are more balanced and conservative than this.

Clearly , sites that produce good quality content ( fresh and authoratitive ), with good quality links ( fresh, authoratative and aged ) , associated with good quality brands, that are talked about ( Enges's 20% ) can rank very quickly now. I saw one new site recently that ranked No1 in a matter of days.

Don't over complicate things IMO - and be patient ( if you have the money and are not having to lay off staff).

But i really think Google could do better to communicate the quality of sites within WMT without compromising their "secret sauce". Not all effected sites deserve to be treated with contempt. Here's hoping.

[indyank]

Google is lying.Obviously lying.Can I show them how many articles are on the same topic with keyword variations from e-how or softonic or any forum or even their own youtube or blogspot platforms.there are so many highly ranked websites that do exactly what they seem to be preaching not to do.But the algo doesn't affect them.

If they aren't lying then their algo is not working against those sites or they deliberately exclude them from their new rules.