Thin content definition and its play in the Pandamonium

[whatson]

Ok, so we all know that duplicate content was a main crack down for Panda, despite it partly backfiring with scrapers out ranking original, etc.

But the other side we hear about is thin content. People saying that every page should have at least 400 or so words of unique content.
How true is this? Is it feasible? Is it fair? Is it relevant?

What about sites like: e.g.
A java game site - the only content is the java app to play a game.
Photo/image sites - the content is all in the images, perhaps their could be some accompanying text, but what if you have 1m+ images.
Flash site - some flash tool that people find useful, interactive map perhaps.
Business directory - like the yellow pages, they have web pages about businesses, but only contain the name, address and phone number.

And many more similar pages, that simply cannot or do not require much text. Are these pages/sites to suffer from Panda? Is there evidence of such sites being Pandalized?

[tedster]

People saying that every page should have at least 400 or so words of unique content.
How true is this? Is it feasible? Is it fair? Is it relevant?

People may be saying that, but Google never said it. And you can easily see examples in the SERPs where it isn't true - you've just mentioned a number of them.

[Leosghost]

I only have 200 words on one of my mini-sites ..and it would probably work with less (around 50 of it's words are TOS ) ..But it is not thin content, nor is it shallow, nor spammy or cheesy..etc.

It doesn't need any more words,IMO, and Google and other SEs apparently agree, and have done so since almost it's first day, it didn't get touched by Panda ..it has stayed at # 1 ( with minor daily "floats " to and "dips" to #2 or#3 from time to time depending on SE's A/B/C testing ) from what is now nearly 500 million results ..nor does it have loads of links..

It is focused and reads that way and looks that way ..on all devices it is very clear what it is about.

It is a KW1,KW2.. 7 letter domain..I have many more like it ..I like KW domains..and "brandable" domains..so do, IME, SEs.

Not smirking ..just saying..so as to prevent FUD from being taken as fact.

[Planet13]

I only have 200 words on one of my mini-sites...

How many pages are those 200 words spread across? Is it just a one page site with 200 words on it?

How is it monetized?

I think these factors might have an equally important influence.

For what it is worth: Many of the pages on my site that are under 200 words do seem to be holding up ok post Panda. they aren't really monetized, just articles on my ecommerce site.

[tedster]

Thin: Reproducing a feed that is commonly used around the web, and adding no extra value

Shallow: Words (any number) that don't actually communicate much.

[wheel]

Ok, so we all know that duplicate content was a main crack down for Panda, despite it partly backfiring with scrapers out ranking original, etc.

But the other side we hear about is thin content. People saying that every page should have at least 400 or so words of unique content.
How true is this? Is it feasible? Is it fair? Is it relevant?

I don't think that's the case.

If you recall, this update was originally called farmers, because it was designed to crack down on content farms. I still think that's a better name for it, it keeps one focused on what the intent was.

[Leosghost]

Is it just a one page site with 200 words on it?

Yes

It offers services ..it says what they are ..it gives an email ( obscured via js ) to contact ..it tells the visitors that the email is obscured by js and that it is safe for them to allow js..it says what the TOS is..

It has no ads..we sell the services ..if we accept to work with the visitor ..it tells them that we may refuse to work with anyone at our choice.

It communicates extremely well..

I have many more one page sites like this..( "description" with TOS and js email )..in many languages ..and for many subjects..but very similar wording..very direct, very clear..they do very well ,the response is good , the page does most of the triage, the services are not cheap.

I can afford to choose to whom I will reply, the services are in great demand, many others offer some of them, few of us actually do them well.

I have some larger sites, and some much larger sites ( not just in English ), both ecommerce, and informational, some of the latter are running ads and adsense ..non of the ads are "in your face"..they do very nicely both in serps and in terms of revenue..

They also as tedster says "communicate well"..they are clear, they look good, I designed and made all of them, and wrote all copy, produced all images and graphics etc and virtually all the photos ( I have bought a few stock photos occasionally, I prefer not to however, I want what I have to be unique ) ..they were made this way so as to be clear and attractive for the visitors...long before Google said that was how they wished sites to be.

I have some which give away things** ..and have ads..and some of those don't have ads

I have some which have info and have no ads..and some do.

I have some which are for fun ..and have ads or not ..

Their "niches", subjects and styles are very varied, one would not think they had the same owner overall..

Above all ..they do not look or feel as if they exist only to show ads or as if the business behind them exists only to push ads on the web..

Because it doesn't ..and I don't.

One can be good at some things ,and successful at them without them being the focus of ones life..

If you can avoid being stressed about what you do ( busy , flashing, ad filled pages are stressful, they look and feel like one is being hustled by "barkers" at carnival sideshows ) that lack of stress will translate to a clear design style, which relaxes the visitor, customer, client etc ..and they respond better, they stay longer, they are more inclined to buy or read, and they are more inclined to look at an ad if they are not feeling like they are being forced to do so..

That said ..well done tasteful flash and animated elements do work in the right place, on the right sites and the right pages ..if they are really good and not merely moving and tacky..

Something netmeg said in another thread yesterday..sums it up extremely well :)

the primary thing you do is provide content that is so compelling/useful/necessary/entertaining/whatever-applies-to-your-niche that they think to themselves "I'm gonna wanna remember this one."

That's usually what I spend most of my time and resources on.

Do that..and keep the message clear, and don't make them feel like you are hustling them..

And make yourself, as much as is possible, independent of SE's ..and spread your risk ..different sites , different niches, different styles..

**giving away things brings people back ..and they tell their friends etc..G works like this.

Oh ..and "we" is because sometimes I hire freelancers to do backend stuff ..and also because I own companies and they are legally "we"..

Some times the personal touch of "I", works better ..depends on circumstances and intended audience

[Planet13]

Thanks Leosghost. I really appreciate the input.

Something netmeg said in another thread yesterday..sums it up extremely well :)

Yeah, I quickly came to the realization long ago that netmeg (along with tedster, goodroi, Robert Charlton, wheel, martinibuster, et al.,) are far smarter than I will ever be... So I tend to pay close attention whenever they write something on the forums.

Thanks again.

[potentialgeek]

> But the other side we hear about is thin content. People saying that every page should have at least 400 or so words of unique content. How true is this? Is it feasible? Is it fair? Is it relevant?

Ezinearticles raised their minimum article length from 250 to 400 words in the wake of Panda. But that number was never provided by Google. It has never offered any minimum word count advice to webmasters. Your guess is as good as mine. In my experience one sentence per page isn't usually enough; whereas at least two paragraphs can do okay.

I don't think it's fair, or even necessary on some sites; but Google doesn't seem to know how to differentiate between legitimate image-rich and text-rich sites, even though it should be relatively easy to engineer.

It could even use the cue of frequency in the number of times "pictures" is included in searches for a keyword to know that users aren't looking for text. Then lower the requirement for minimum text quantity per page to nearly zero. Users may want a short description and perhaps a comment. Maybe less.

> Is there evidence of such sites being Pandalized?

My first site to get Pandalized was a popular site with art pictures created as a hobby over ten years ago. Many pages were thin and the galleries had been created or autogenerated with photo gallery software. You put all the images in a folder, choose page title and picture description, and then run the software. It cranks out pages with one image per page and the same page title, e.g., Red Widgets Picture (1/50)... Red Widgets Picture (2/50)...

So what I did was add a meta tag to all pages with similar, thin content to de-index them from Google. The rest of the site has unique text. Not sure if it will recover from Panda, but it was worth a shot. Hopefully it helps until Google's engineers become more sophisticated.

[edited by: potentialgeek at 6:39 pm (utc) on Jun 24, 2011]

[hairresources]

When it comes to duplicate content issues addressed in the Panda update, what are the recommendations if you use an Ezine article on one of your webpages.

Wouldn't that be considered duplicate content? Should you noindex these pages?

[tedster]

I certainly would. Then again, I don't work with any sites that turn to ezine for content.

I'm a dyed-in-the-wool believer in original content. I think article resource sites were the first step on the road to content farms. They saw that the "crappy content" model was working and decided to bypass the middle man and source their crap more directly.

[caran1]

Ultimately "good content" should provide the information that the visitor to the website is looking for.
Whether it is split into one page or ten pages, it is entirely upto the website owner , who pays all the website costs like hosting and domain registration.

[tangor]

Ultimately "good content" should provide the information that the visitor to the website is looking for. Whether it is split into one page or ten pages, it is entirely upto the website owner , who pays all the website costs like hosting and domain registration.

I've experimented across 10 offerings on the web, some with pagination, some with "bitty" content" and the rest with full-strength regardless of length articles. Two paradigms to consider: frequency of access and time on site. My "cost" is expressed in bytes and 10 pages to produce one article is greater than one article (includes headers, images, footers, etc. and don't tell me that's not being done! And at a bandwidth cost, too.).

Pick your poison... both pagination and full-strength sites seemed equal in conversion rates (Conversions: 22.2% or Time on site: 23.9%), but the "bitty" dismal content did not fare well (in the .0 numbers)

[mrmobility]

When it comes to duplicate content issues addressed in the Panda update, what are the recommendations if you use an Ezine article on one of your webpages.

Wouldn't that be considered duplicate content? Should you noindex these pages?

Surely that is a perfect case for a rel canonical? Just because you aren't the original source doesn't mean you shouldn't give Google the opportunity to rank you if it thinks you are the best result for that article.

[caran1]

@tangor
If depends on how your website is designed, if you are using only a small fraction of the bandwidth/resources available to you, the additional bandwidth or CPU usage does not matter .

[tangor]

If depends on how your website is designed, if you are using only a small fraction of the bandwidth/resources available to you, the additional bandwidth or CPU usage does not matter .

It all matters, all the time. If one takes a different approach, well, expect costs to evnetually creep because the "bad boys" will only increase... they will never get "less".

[whatson]

I think you know when the content is sufficient when you truly believe that is good enough to click that FB Like button, and actually want to share it with others. This is the level of quality you want to aim for.

[potentialgeek]

I found a webmaster who claims his site was unPandalized after he developed content:

"I did enough to recover, but I still have some work to do and it takes time. Here's what I did: Removed posts with pictures and little text. Removed posts with videos and little text. Removed posts less than 150 words, except for the ones I rewrote to make them over 500 words... You don't have to write any essays, but 150 words is simply not enough. I shoot for 500+, but I'm not concerned once I go past 300 (and it's really easy to go past 300 when you start writing something worth writing). [Note: I use 150 as the marker because this is the size of most excerpts, which is what the scraper sites tend to use.]"

I don't believe there's a fixed number minimum word count for Panda, but the more I research, the more it seems page length makes a difference. My sites which were Pandalized were undeveloped with short pages. The sites which weren't Pandalized were developed with longer pages. The pages which still get good rankings on Pandalized sites aren't short pages. One of my pages that recovered from Panda also isn't a short page.

Page length isn't the only issue with Panda, but it's a major issue on the road to recovery.

[hairresources]

I am finding more and more a major difference in ranking almost entirely based on social links, i.e. facebook, twitter mentions- everything else being equal. I would rank the number 1 and 2 problems for pandalized sites to be 1. thin content and 2. lack of social connections

[Whitey]

To put this another way: I have never ever seen a site where everything was great and it would have ranked well except its articles were only 200 words long. That's just... not the way the algorithm looks at stuff. If your site isn't ranking well, the cause must be elsewhere Susan Moskwa google employee : [google.com...]
[webmasterworld.com...]
Look at how you can add *real* value to the user experience compared to others, not the length of the content, or, content that says the same thing a different way. There's unique content and there's content with a unique experience ... if you get my drift on the difference.

[freejung]

potentialgeek, while I do think you're on to something, there seems to be a tradition that whenever anyone makes a statement about Panda, someone comes up with a counterexample. So here's one:

I have one page on my pandalized site that was not affected at all. The page has about the same amount of text content as the pandalized pages, less than many of them. The subpages (individual image pages) actually have quite a bit less than average, the worst offender being about 5 words. So overall the text content of the category page and its children is quite a bit less than average for the site.

The difference is, this one category is amazingly popular (as measured by various metrics, including Facebook likes). Much more so than any other category on the site. Clearly its obvious popularity is what made it immune, despite its relative lack of text.

My point is that while in some cases text length may be an important issue, if a page is sufficiently popular it can have less text and still be fine. Likewise I would expect that if a page is sufficiently unpopular it will be in trouble even if it has quite a bit more text.

Edit to add:

I would consider the possibility that the correlation you're seeing with text length is a secondary effect. The longer articles may have covered their subject in more depth with more useful information, thus creating a better user experience and causing those pages to do better. To put it another way, I doubt you could have achieved the same effect by simply padding the pages with nonsense. In the course of adding more content, you also added more value, thus improving the status with Panda.

[Rasputin]

I would consider the possibility that the correlation you're seeing with text length is a secondary effect. The longer articles may have covered their subject in more depth with more useful information, thus creating a better user experience and causing those pages to do better.

IMO this is absolutely correct and key to understanding panda, which I believe is essentially based on 'the long click' (where people find what they are looking for and don't go back to look at the next result)

So a 20 word article can be plenty if you search for 'how old is the queen' or 'what year did WW2 start'.

Pre-panda a decent site could get away with adding lower quality (thin content) articles and they would rank because of the overall site ranking. So many people (including me) did exactly that.

Post-panda the situation is exactly reversed and now the whole site is demoted precisely because of those lower quality articles.

(Note: this 'insight' hasn't got my main site out of panda yet but Ive got quite a lot of rewriting to get through!)

[Hissingsid]

@Leosghost I found your post here very interesting. I have a few smaller sites with unique content. Some I did a few years ago and have basically left well alone, others I make additions to on rare occasions. All of them continue to rank very well for the terms they target.

I'm interested in whether, for your one page sites, do you tweak them regularly or just get them near perfect for your purposes and leave well alone?

Cheers

Sid

[sundaridevi]

Reading this thread, two thoughts came to mine. Firstly, the "Google Quality Raters Handbook" specifically makes a distinction between different roles for different kinds of pages. It goes on to say that google tries to determine the user intent by the words used in the query. So if your page is a result for what the call an information query (e.g. what is chocolate?), probably more words are better. They divide queries into different categories. I suppose some of them require more info and some less.

My 2nd thought is more relevant I think. I've long thought that google favors very long pages and my best performing pages have been those (e.g. one page ranked number one for ten years for one of the most competitive city search terms). There are two onsite factors, that most SEOs will tell you don't matter, that matter immensely if you take word count into play. While those factors have never been the most important (in google) they did matter if you were in a very competitive field and wanted to move from say number 5 to number 1. So even pre-Panda, word count was important.

That's why I don't like this term, "thin content". I think you provide the content to get the job done. 500 words of junk is just as thin as 100 words of junk. If nobody links to a 1000 word page you won't do well. If everybody links to a 50 word page you will do well. All else being equal, maybe post-Panda 500 words beats 50 words.

[tedster]

That's why I don't like this term, "thin content".

Agreed. To Google, the phrase "thin content" means something very specific - a website that does little beyond publishing an affiliate feed in their own template. It is not a phrase addressed to content length. It is not the same thing as what they called "shallow content" when Panda rolled out. Shallow content can also be any length - but when you finish reading, it contained very little or no real information -- just enough not to trigger the spam filter.

These are Google's own descriptions. If webmasters allow the meanings to get crossed up, they will end up chasing their own tails with website changes that do not have any chance of fixing a Panda problem.

[potentialgeek]

See this page for new data on Panda:

[searchenginewatch.com...]

[Leosghost]

@hissingsid

just get them near perfect for your purposes and leave well alone?

I build them on the server(s),takes an hour or two max, when I like how they look and read, I look to see if there is anything I could remove to make them better..

just get them perfect for your purposes and leave well alone?

:)

Then make 'em live, and move on, one gets to "grok" what will work, and I still have hundreds of domain names to take "live"..

[Hissingsid]

@Leosghost

Tantalising. You've inspired me to do some experiments with some of my dormant domains.

Many thanks

Sid

[indyank]

potentialgeek, that was a good read.

But I am not so sure about the usefulness of those tests. Those folks seem to have assumed they knew what google considered as "low", "medium" and "high" quality, but failed to define or give examples for those content types. If people knew that, there would have been many more recoveries.

It would have been far better if they gave out examples of sites for the various types including the ones that recovered. The tests would have been better appreciated if they had given examples of a site that recovered, how it looked before recovery, the kinds of pages they considered to be of low quality in it, and how it was improved to what it is now, to recover.

The conclusions were interesting but for the last one.

You can recover from Panda by removing low quality pages from the index (canonical, noindex, etc.)

I haven't heard anyone say that they recovered from panda by adding "noindex" to what they considered as "low quality" pages.

[MrSavage]

I think overall people are missing the point of thin content. I believe at this point that thin content is relative. In other words, monitization vs site or page content. This is preliminary, but if you ask me today, I think for each ad, affiliate link, etc, it chips away at your content value. I'm still working on this. However, what does every content farmer have in common? Making money. If they aren't making money, then their business model ceases to make sense. So for me moving forward, my way out of Panda will be based on content vs. monetization.