Articles penalised for being too long?

[Rasputin]

I have one site which I haven't looked at much for a couple of years but have just started improving the site, mostly by thinking about rewriting 'thin' articles.

But it is clear after investigating that the articles on our site that are performing the worst in the serps are not at all thin - in fact they tend to be very substantial articles, typically 1000-2000 words whereas the articles in the serps including wikipedia are only 100-200 words.

Our articles are not fluff and padding, but relate to a widget where we have an expert in the field writing about them, and elaborating far more than other sites do (the author is a university professor on the subject).

To confuse matters slightly, the articles relate to a 'foreign widget'. The serps in the original language (Italian) are as I would expect, long and detailed articles like our own (including wikipedias), but these more detailed articles don't exist on English language sites, except on our own.

Our articles that are less detailed and of similar length to articles on the same subject on other sites seem to rate much better in the serps. So it seems reasonable to conclude that if an article is much longer than some 'google average' for that keyword then it is moved down in the results.

Could this be correct? If an article is much longer than the competition do google automatically assume it must be padded or spam? If so what can I do about it?

[goodroi]

author is a university professor on the subject

You might want to double check the readability of the article. Being too smart can actually make your content less approachable for users. Each industry & keyword is different but generally I try to balance the information value of my content with its ease to be digested by readers.

[Rasputin]

This is definitely a consideration, but 'luckily' his English is not great so I already need to rewrite them to be clear to an English language audience.
Double checking that the language is suitably user-friendly (easy to read) is part of what I am currently working on.

[goodroi]

You might also want to look at keyword density. Before people get the wrong idea I do not think keyword density has a significant impact on rankings. I do like keyword density to quickly breakdown a page (especially a 2000+ word article) so I can see what phrases are being used and possibly what keywords are being overlooked. You might find too many scientific or technical terms to make it attractive to the common user. It is quick and easy and gives a fresh perspective on your content but do not overvalue the importance of keyword density.

[EditorialGuy]

So it seems reasonable to conclude that if an article is much longer than some 'google average' for that keyword then it is moved down in the results.

That hasn't been my experience. Our site's highest-ranking articles tend to be long, not short (and they certainly aren't in the 100- to 200-word range).

As always, YMMV.

[netmeg]

If I'm not sure whether long blocks of content are maybe too wordy for readability, I close the door and read them out loud to myself. If they don't make sense or come across as stiff or unnatural when spoken aloud, they're probably not great as written content either.

[Planet13]

Just a thought...

depending on the audience, over-simplifying can be worse that under-simplifying.

[Rasputin]

I don't think there is a problem with readability etc. I'll try to explain better without bring specific or mentioning keywords etc...
Imagine a small medieval hilltop village somewhere in europe, lets call it 'anyoldplace'. The place is so small and out of the way that most guides if they mention it at all might say:
'Anyoldplace is a pleasant village to explore, with cobbled streets leading up the hill to an ancient church'.
An enthusiast of architecture, the region and art history could easily write 1000 words about the design and style of the church, the scenery and places of interest close by, frescoes in the church, small architectural highlights in the village that a casual observer might not notice, the local traditions and cuisine...and much more besides.
This could of course be a well written and interesting article, but is the type of article that is failing to rank when written in English - although the equivalent size and type articles on other sites rank well in Italian search. So my query is whether it is the relative length of the article compared to competitors that is the problem, rather than whether it is a problem with writing style, correct use of English etc.

I have given no thought to keyword density but it would tend to be quite low (title, a couple of headings/ subheadings and once every 2 or 3 paragraphs, I think)

[superclown2]

Searchers are impatient. They haven't time to read through reams of text when the next site is only a click away. They want information to be easy to find, in bite-sized chunks. Try breaking the text up with suitable headings to make it easy to navigate.

[Rasputin]

Thanks for all the comments. I'm not quite sure what I have said that gives the impression that it is a problem with the readability of the text.

The articles are very readable, and actually quite interesting. When visitors do manage to find them the average time on the pages is good, typically more than 5 minutes, which is certainly better than most of my other sites.

The problem I have is getting these pages to show in the serps, rather than making them decent quality in the first place...

Thanks

[netmeg]

They may well be very readable and useful and interesting, but still be too long for average readers (and if Google detects that, it could affect where you shake out in the SERPs).

If it were me, I'd test displaying them in different ways, to see if that helped any.

[EditorialGuy]

Searchers are impatient. They haven't time to read through reams of text when the next site is only a click away. They want information to be easy to find, in bite-sized chunks.

Depends on the searcher, the topic, and how interested the searcher is getting in-depth information about that topic. Not everyone is looking for Cliff's Notes: On our site, significant percentages of pageviews are from readers who stay 30+ minutes or visit 20+ pages in a session.

[goodroi]

You could head over to mturk and get a bunch of people to review a few different versions to find which one is best.

[Planet13]

"...whereas the articles in the serps including wikipedia are only 100-200 words."

I think the word count might be a red herring.

I doubt that the algo looks at a subject and says: "articles with only 200 words are BETTER than articles with 2,000 words for this subject / keyword."

[Saffron]

My longest (and by far the most popular article on my site) was penalised by Google last year. Pushed down from no. 1 to no. 6, with much thinner articles replacing mine.

I think it depends on the topic too. If you're researching 'should tomato sauce be stored in the fridge or pantry' then of course you don't want to wade through 2,000 words of text. Other topics (I can't think of any off the top of my head, just woken up) do require much more detail.

There is one site I quite like but the author does go on and on and on a bit, and adds stupid and unnecessary images to his articles. I can see he's quite a good writer, I just wish he'd cut out the fluff, he could reduce the content by 30% and probably be much more readable.

[EditorialGuy]

One thing to keep in mind: There are zillions of sites cranking out short, shallow 200- to 500-word blog posts, articles, etc. (often by people who have no firsthand knowledge of the topics being covered), so sites that have articles with more substance have a unique selling proposition that the content farms (and, these days, "content marketers") lack.

I'd rather stand out from the crowd than be part of the crowd.

[superclown2]

sites that have articles with more substance have a unique selling proposition

Substance and words are two entirely different things. Many authors are, unfortunately, paid by the word and it shows.

[edited by: superclown2 at 10:24 pm (utc) on Jul 15, 2014]

[superclown2]

On our site, significant percentages of pageviews are from readers who stay 30+ minutes or visit 20+ pages in a session

On my sites they could get a quote 30 times in that time %-)

I'm happy if they stay 3 minutes and buy the product. 30 minutes means that either there's a h*ll of a lot of info or it's impenetrable. Personally - for what it's worth - I would break it up over several pages, perhaps using hyperlinked bullet points. That way each page would be that bit more relevant for the appropriate search term and more likely to have a decent position in the SERPs for those search terms.

Ten tightly focussed pages beat one more general one any day of the week.

[austtr]

An enthusiast of architecture, the region and art history could easily write 1000 words about the design and style of the church, the scenery and places of interest close by, frescoes in the church, small architectural highlights in the village that a casual observer might not notice, the local traditions and cuisine...and much more besides.

I realize that is just an example to help explain the OP but it does highlight an issue that I suspect hinders long pages of content.

Within a page that you describe, there is going to be content about:

the town
the region
the church
The local art history
the architectural highlights
the local traditions
the local attractions
the local cuisine
the local transport
etc

When you write a page that covers all of that, and I'd expect around 2000 words minimum, then not only is it a long page but there is no single, over-riding topic that is dominant throughout the content.... by their very nature, these types of pages have fragmented topics.

It could well be that any ranking issues are more about SE's being confused by the fragmentation. I would not expect a long page to suffer purely because of its length.

As a writer of destination guides, I have struggled to come to terms with this issue many, many times. Do I leave my 2000-3000 word, all encompassing pages written for the viewer... or... do I break up those long pages into 4 or 5 smaller, more focused pages so that the SE's can better understand the topic? And create a navigational nightmare in the process?

[Planet13]

Firstly, I apologize in advance if my suggestions aren't helpful. I personally appreciate what you are trying to do as I disdain those hit-and-run articles that rank well for specific keywords.

" or... do I break up those long pages into 4 or 5 smaller, more focused pages so that the SE's can better understand the topic? "

Is there a way to break up the long pages so that they are better for both SE ranking AND the reader?

I know there were some discussions about using the next - previous tags, and I remember tedster (RIP) wrote that great post here about long form journalism.

And also I know there are more than a few tech sites out there that post a single article across two or three pages.

Anyway, I hope this helps.

[EditorialGuy]

Superclown2, this thread is about articles, not sell pages. No one is suggesting that a widget sell page should be the equivalent of an article in THE NEW YORKER.

Austrr wrote:

As a writer of destination guides, I have struggled to come to terms with this issue many, many times. Do I leave my 2000-3000 word, all encompassing pages written for the viewer... or... do I break up those long pages into 4 or 5 smaller, more focused pages so that the SE's can better understand the topic? And create a navigational nightmare in the process?

If it's logical to break a long article into separate pages, then do it. There's no reason a multi-page article has to be a "navigational nightmare." In the case of an in-depth destination guide, it would make perfect sense to have dedicated pages such as:

- Introduction to Widgetville
- Sightseeing overview
- Museums
- Hotels
- Restaurants
- Transportation
- Tourist offices and other resources

Have an internal menu for the article that makes everything easy to find, and the reader will be happy.

I'd also link the pages together with link rel="prev" and link rel="next" so the search engines will understand that the pages, while focused, are part of a single entity. (Schema markup can be useful, too.)

[Planet13]

+1 editorialguy

Also, I would suggest to look at the devices / screen resolution of your users and see whether you might need to make changes to accompany them.

[tangor]

We've had a conversation regarding readability. See: [webmasterworld.com...]

Some useful info there that might help with understanding why some longer articles may not fair well.

[Rasputin]

Just to reiterate:
the articles I have that are long and where competitor articles are also long do very well, some of our best performing pages.
The articles I have that are long but where competitor articles are very short are not doing at all well

This is what makes me wonder if, for a given search term, there is some kind of 'expected length' with results much shorter or much longer being excluded or demoted in the results.

This would probably make sense because as pointed out it certainly isn't always the case that more words is better.

[tangor]

More likely it is the quality, or the perceived quality, of the articles in question. G and B do rate everything for readability, which is part of the 200+ items used to rank returns.

[bumpski]

To confuse matters slightly, the articles relate to a 'foreign widget'.

I'd propose a completely different direction of research; Google has not only discussed quality of pages, but localization of sites and pages as well. Google's algorithm may be incorrectly localizing these long articles to a certain region of the world, only providing higher rankings when the search is done from the region that the algo has determined the page was targeted at by the author.

Regarding quality, one could compare the reported reading levels for the shorter and long pages, not that the reading level should have much to do with quality. Use "search tools", "all results", "reading level".

Google Analytics may be able to help one find localization problems, just remember, if it's smaller sites, use a regional custom filter to remove your own activities on the site. Here in the US my ISP, moves me from city, to city, to city, for each new IP address, assigned. (I'm mostly glad I don't have a static IP). I believe some significantly large sites may have been incorrectly localized by Google's algorithm.

[EditorialGuy]

Also, I would suggest to look at the devices / screen resolution of your users and see whether you might need to make changes to accompany them.

One of my beefs about "responsive layouts" is that simply shifting blocks of text around on a page doesn't make the page mobile-friendly. Often you need to break paragraphs into very short units (a sentence or two, in some cases) to make them easy on the eyes if the reader is using an iPhone or another small-screen device. At the same time, if you take a "mobile first" editorial approach, your pages may look awful (or seem skimpy) on tablets, laptops, and desktops.

[Planet13]

"If an article is much longer than the competition do google automatically assume it must be padded or spam? If so what can I do about it?"

I still don't see this as being likely, but it is worth experimenting.

It could be that - based on user activity - there are signals to google that users prefer shorter articles on certain subject.

The thing is, I haven't seen this as being a universal truth though. Some articles I have are short and they out rank a lot of longer articles, some articles are long and outrank shorter competition.

~~~~~

It sounds like your search term is maybe two words (or more). What happens if you flip the order of the keywords? (Even if it isn't commonly said that way.) Are the results the same?

[Rasputin]

I like the idea about checking reading ability, although it looks like both our performing and non performing pages are ' intermediate' as are most of the other pages listed (with a few 'basic' level pages in the results as well, but they are mostly brand sites turning out stuff I wouldn't aim to duplicate).

Re keywords, these are such unusual and unsearched for widgets with no sensible competition that I would expect to see them with 1 word search. They rank better for longer phrases.

Their obscurity also means they will always be low visitor number pages as well, so won't ever make any money, but I'd still like to understand why they don't rank and hopefuly learn something that can be applied to other sites.

Re mobile / desktop our template is responsive and looks ok on most devices. The need to look ok on mobile means paragraphs are rarely more than a couple of sentences long. Most visitors use tablets / desktops, perhaps due to users generally not being particularly young.

I'm getting the idea from all the responses that the problem is more likely to be some other signal from the pages rather than a simple question of page length so that's what I'll focus on: it is certainly possible that users spend a long time on the page but still go back to the serps to find something quicker to absorb.

[EditorialGuy]

"If an article is much longer than the competition do google automatically assume it must be padded or spam?

Not in my experience.

One thing to keep in mind, though: Writing long, in-depth articles (and knowing how to make them enticing to readers) isn't a job for amateurs. It's harder than it looks.