PDA

View Full Version : Auto translated pages vs google content guidelines



StarBuG
30-03-11, 20:20
---
ADMINISTRATOR NOTE: this post includes false information about Google statement for automatically translated content. Also the statement do not applies to vBET. Please see here for details: http://www.vbenterprisetranslator.com/forum/vbet4-general-discussions/1612-auto-translated-pages-vs-google-content-guidelines.html#post7873
Unfortunately I was on vacation when message was created and lot of wrong information was included in this thread - this is why correction is included here.
---

Hello

I just wanted to let you know that Matt Cutts said that auto translating webpages falls within their "auto generated content" guidelines which are punished since the Google Panda update.

So just as a warning you might want to reconsider adding all your translated pages into your sitemap.

In fact you might want to consider to block google from listing auto translated languages at all via robots.txt.

Here is the video: YouTube - How should I handle localized content? (http://www.youtube.com/watch?v=GyWx31GeQWY)

Regards

StarBuG

norku
30-03-11, 21:49
Oops... seems pretty clear that Google doesn't like what Vbet does then. Why does Google still index auto-translated pages though? Doesn't make sense.

StarBuG
30-03-11, 22:16
Because when it visits the pages it does not know that they are auto translated.
They are stored in the database and are not served via their api to their crawler.

It is totally fine what vBET does however not in the sense most people try to use it.

It is a translation feature that helps people speaking different languages to understand your content.
However it is not a way to multiply your page content.
If you use it that way Google will see this as low quality "auto generated content" and punish your website for it sooner or later.

norku
31-03-11, 08:48
Thanks.

Kamil,

Can I just confirm please that if I want to keep Vbet running on my forum but do not want Google to see the auto-translated pages that I need only to do the following:

1. Enter 'Disallow: /forum/xx/' into robots.txt where xx denotes the foreign language(s) in use on my forum

2. Reverse the steps in this post for sitemap integration: http://www.vbenterprisetranslator.com/forum/vbet4-integration-other-plugins/415-sitemap-generator-integration.html

3. Re-generate and re-submit sitemap.

kamilkurczak
31-03-11, 12:34
yes. if you disallow your pages for google bots - you will have your pages not indexed by google.
of course if you reverse all changes from sitemap - the same - page not indexed.

of course already you have the pages indexed by google you you need to reload the new sitemap.
****
Added by Michał Podbielski:
You do not have to block any translated content, see:
http://www.vbenterprisetranslator.com/forum/vbet4-general-discussions/1612-auto-translated-pages-vs-google-content-guidelines.html#post7873

norku
31-03-11, 13:15
Thank you Kamil. I think it's a real shame that Google doesn't like content generated in this way. I think they should make an exception for forums because Vbet is a fantastic idea and even in the short time I had it fully working I saw an improvement in my stats. All is not lost though of course because the really clever part of Vbet - that foreign language speakers can participate on a forum in their own language - is of course still there.

tavenger5
31-03-11, 14:00
Hrm, is google eventually going to just drop those translated pages? If that happens I wonder if that will affect the ranking of the native language pages?

MOGmartin
07-04-11, 11:31
Hrm, is google eventually going to just drop those translated pages? If that happens I wonder if that will affect the ranking of the native language pages?

Google do appear to have removed VBET pages now: please see the discussion on my blog.

I personally DO NOT THINK that these pages are spammy, they have a legitimate purpose...
Lost in Translation: Google Algo Update 06-04-2010 | SEO Blog from SEOforums.org (http://seoforums.org/seo-blog/google-translation-algo-update-06-04-2010/)

vBET
07-04-11, 14:41
Hello

I just wanted to let you know that Matt Cutts said that auto translating webpages falls within their "auto generated content" guidelines which are punished since the Google Panda update.

...

Here is the video: YouTube - How should I handle localized content? (http://www.youtube.com/watch?v=GyWx31GeQWY)


That is NOT true. Please see the video once again. He said that it CAN apply to auto generated content guidelines and then he is telling about wrong impression when page is totally not readable for humans. So really Google robots are not able to consider such pages as auto generated content and this could be done only by human and only if translation is even not week but totally crashed, because having translated pages itself is very good what is described in the movie on context how URL's should look then.

Also I want to calm all our users, that what said Matt Cutts about automatic translations which CAN falls Google's auto generated content guidelines is NOT APPLICABLE to vBET. It is because:
1. Matt clearly says that he is talking about simple translation of whole content, and vBET do not translates whole content - it saves user names and other areas if marked as not for translation for vBET.
2. vBET is not about automatic translation but about supporting multi-lingual communities, where everyone can write in its own language in same thread. So once again it is not just simple translation of whole page, and it is not about automatic translation but about multi-lingual communities.
3. Translations itself are made by Google itself which has very good quality and as Matt clearly said such translations CAN be treated as auto generated content only if those are really not human readable.

So hope our users feel better now :) To avoid bad feelings and scaring those who are not reading whole topics just headers I will change title of this topic. With full respect to the author - we are doing this only to avoid misunderstanding, because topic assumes something very bad which we believe is not true. Hope it is OK for author :)

vBET
07-04-11, 14:47
Thanks.

Kamil,

Can I just confirm please that if I want to keep Vbet running on my forum but do not want Google to see the auto-translated pages that I need only to do the following:

1. Enter 'Disallow: /forum/xx/' into robots.txt where xx denotes the foreign language(s) in use on my forum

2. Reverse the steps in this post for sitemap integration: http://www.vbenterprisetranslator.com/forum/vbet4-integration-other-plugins/415-sitemap-generator-integration.html

3. Re-generate and re-submit sitemap.

You do NOT have to block any of translated content! Please see the movie again - this whole movie is about how to name URL's when you have translated content, so having translated content is perfectly fine for Google!

The note about automatically translated content DO NOT apply to vBET and also StarBuG write false information that it is considered as content duplication what is not true (it CAN be, not "it is", and this CAN do not apply to vBET) - see my previous post.

vBET
07-04-11, 14:52
Thank you Kamil. I think it's a real shame that Google doesn't like content generated in this way. I think they should make an exception for forums because Vbet is a fantastic idea and even in the short time I had it fully working I saw an improvement in my stats. All is not lost though of course because the really clever part of Vbet - that foreign language speakers can participate on a forum in their own language - is of course still there.

This is not true that Google doesn't like such content. This is all misunderstanding because of false information which StarBuG put here. Please look and hear carefully the movie. Here you have it described more:
http://www.vbenterprisetranslator.com/forum/vbet4-general-discussions/1612-auto-translated-pages-vs-google-content-guidelines.html#post7873

vBET
07-04-11, 15:05
Google do appear to have removed VBET pages now: please see the discussion on my blog.

I personally DO NOT THINK that these pages are spammy, they have a legitimate purpose...
Lost in Translation: Google Algo Update 06-04-2010 | SEO Blog from SEOforums.org (http://seoforums.org/seo-blog/google-translation-algo-update-06-04-2010/)

I read your blog. As I see it is reaction to last day information. Please note that In my opinion it is not related to translations. Especially that I just checked and I see that your forum is indexed in other languages: site:SEO Forum + SEO Blog SEO & Tools@seoforums.org (http://seoforums.org/de/)

So because I see that your translated pages are indexed I suppose that it was only some temporary issue maybe even on Google analytics side, but definitively not related to vBET. Can you describe what happens now?

MOGmartin
07-04-11, 17:08
mate - Im a big fan of your plugin, and I've made the case with google that its not auto translate spam.

Having said that - take a look at my blog post (http://seoforums.org/seo-blog/google-translation-algo-update-06-04-2010/).

The VBET auto translated pages have been booted out of the index buddy.

My suggestion: use a different API or at least give the option to. There is no way that google can detect that pages are auto translated IF THE TRANSLATION IS GOOD ENOUGH.

Certain languages like Spanish work very well, some dont.

MOGmartin
07-04-11, 19:27
I read your blog. As I see it is reaction to last day information. Please note that In my opinion it is not related to translations. Especially that I just checked and I see that your forum is indexed in other languages: site:SEO Forum + SEO Blog SEO & Tools@seoforums.org (http://seoforums.org/de/)

So because I see that your translated pages are indexed I suppose that it was only some temporary issue maybe even on Google analytics side, but definitively not related to vBET. Can you describe what happens now?

the pages are still indexed, but they do not appear for any searches.

Its 100% NOT a temporary GA issue, and the non translated (ie. normal) pages have not lost any search volume.

(almost) 100% of the search volume going to the VBET pages has stopped.


Don't get me wrong - Im 100% on your side here, the issue is, google isnt.

Also - I'm NOT saying that the VBET plugin is at fault or broken - it still works very well and is a great plugin. Google however have announced that they are treating these pages as autogenerated, and as such are not sending any search volume.

I will update you and my blog in a couple of days time.


Some credentials for you: Im a professional SEO, have been for 6 years, I speak regularly at conferences - I'm not someone who is talking with "half knowledges" or a misunderstanding of how google works.

MOGmartin
08-04-11, 11:54
Here is the updated traffic stats.

@hilinebuilders here is the updated graph. Also seen the same... on Twitpic (http://twitpic.com/4i8zwr)3


FYI: I also have sites powered by wp-robot in wordpress that use the google translate API. They have also been hit at the same time and lost 100% of their traffic.

norku
08-04-11, 16:20
Thanks for the new info chaps. I blocked my translated pages on the 30th of March and then decided to reverse the changes a couple of days later so I've had Vbet running normally again for a week now. I just checked my landing pages and pageviews and can see no significant drop in activity for the Vbet-generated content. Things seem normal here but I'll monitor the stats and if I see the Vbet pages being dropped I'll post here.

MOGmartin
08-04-11, 16:28
Thanks for the new info chaps. I blocked my translated pages on the 30th of March and then decided to reverse the changes a couple of days later so I've had Vbet running normally again for a week now. I just checked my landing pages and pageviews and can see no significant drop in activity for the Vbet-generated content. Things seem normal here but I'll monitor the stats and if I see the Vbet pages being dropped I'll post here.

its not the activity (ie. pageviews) that you should look at, its the landing pages, ie traffic that google send to those pages.

the activity is perfect, the VBET plugin is perfect, the issue is google not sending traffic to the pages it creates (which is a big part of the plugin's value to me)

norku
08-04-11, 17:19
As I said, I checked the landing pages and I see no significant difference. Mine is a small forum and I've only been using vbet for a few weeks but my results aren't showing a drop at the moment.

Edit: I can't upload pics here but I am looking at the graph here (filtered as yours is) and it shows a steady climb. Perhaps I'll see the drop on the next update but certainly for now I see no probs.

vBET
09-04-11, 19:54
mate - Im a big fan of your plugin, and I've made the case with google that its not auto translate spam.

Having said that - take a look at my blog post (http://seoforums.org/seo-blog/google-translation-algo-update-06-04-2010/).

The VBET auto translated pages have been booted out of the index buddy.

My suggestion: use a different API or at least give the option to. There is no way that google can detect that pages are auto translated IF THE TRANSLATION IS GOOD ENOUGH.

Certain languages like Spanish work very well, some dont.

Just as I wrote I read your blog already :) And I wrote you that I see your translated pages by vBET indexed in googe. See yourself - just write in google query:
site:http://seoforums.org/de/
So those are not booted out of the index, or maybe I do not understand what do you mean...

vBET
09-04-11, 20:21
the pages are still indexed, but they do not appear for any searches.


This is not true. I just send this query to Google:

wenn ich tat Blog-Posting zu diesem Zeitpunkt eine Sache in meinem Kopf kommen. Was ist Haupt-Unterschied zwischen Autoblog und spamblog
The result is only one and it is your site on page translated by vBET.


Its 100% NOT a temporary GA issue, and the non translated (ie. normal) pages have not lost any search volume.

(almost) 100% of the search volume going to the VBET pages has stopped.


Don't get me wrong - Im 100% on your side here, the issue is, google isnt.

I'm glad you like vBET :) This is not about who is on which side. We just must check any information which can be crucial for our users. Especially if it is information so important like this. So we just have to correct any statement which could give false impression of vBET users about its usability.

That is why we checked does it is true that your translated pages are removed from the index. And now we checked does it is not possible to reach those by searching. And in booth cases results are good - it means that your pages are indexed and are reachable by searching.

Please note that if your traffic for translated pages is lower that before it could be good to compare your positions in Google index, what is very important for traffic from searches. Maybe because of Google changes some of your pages wend down in index. It could be temporary or for longer. Hard to tell.


Also - I'm NOT saying that the VBET plugin is at fault or broken - it still works very well and is a great plugin. Google however have announced that they are treating these pages as autogenerated, and as such are not sending any search volume.

I will update you and my blog in a couple of days time.

Thanks for note. Of course we also have to check such information analyzing does it is vBET fault or not. This is to assure great product for our users and as one of our users we hope that you are glad that we treat it seriously and checking everything also from side of vBET potential bugs.

We will wait for your update.



Some credentials for you: Im a professional SEO, have been for 6 years, I speak regularly at conferences - I'm not someone who is talking with "half knowledges" or a misunderstanding of how google works.

We never suggest that you do not know what you are writing about. And really sorry if you took such impression. Please note that this is not private discussion but public board so we MUST include all crucial information even if we think that you know about this, because it is also for whole our society. So if you see some obvious information for you, it is just because we care about all our users to understand more this topic :) Hope it is OK for you - if not you can switch to private discussion with our staff :)

vBET
09-04-11, 20:37
Here is the updated traffic stats.

@hilinebuilders here is the updated graph. Also seen the same... on Twitpic (http://twitpic.com/4i8zwr)3


FYI: I also have sites powered by wp-robot in wordpress that use the google translate API. They have also been hit at the same time and lost 100% of their traffic.

Kind of stupid question - where you can put this filters in analytics? I just made fast check and didn't find it - you already know where so I would like to safe some time for looking :)

We are on it and checking it. It seems to be individual - some forums see such tendency some not. At this time we are not able to determine what is the reason. Just in case for enyone who has such issue - please make sure that you didn't block all translated pages in robots.txt

MOGmartin
11-04-11, 11:57
Kind of stupid question - where you can put this filters in analytics?

no probs michal,

its: CONTENT -> Top Landing pages
FILTER: Pages containing this string:

/bg/|/hr/|/cs/|/da/|/nl/|/fi/|/fr/|/de/|/el/|/hi/|/hu/|/it/|/no/|/pl/|/pt/|/ro/|/ru/|/sr/|/sk/|/es/|/sv/|/th/|/tr/|/uk/

Updated picture with recent stats, and you can see in the left column where the report is:

Twitpic - Share photos and videos on Twitter (http://twitpic.com/4jl0oo)


please make sure that you didn't block all translated pages in robots.txt

FYI - I didn't block traffic in any way ;)

JUST TO RE STATE:
I do NOT beleive that the vbet plugin is broken. It still does EXACTLY what its advertised as - people can still use the translated pages 100%.

The ONLY issue is that google have stopped sending traffic to these translated pages. As its damn hard to footprint a translated page if its a good translation, I suggested that you incorporate the yahoo translate API.

I would bet good money that this will make it much harder for google to filter out of the index.

JUST TO RE STATE AS WELL:
I am NOT SAYING THESE PAGES ARE DE-INDEXED.... They just are not returned in the search results when they used to be.

tavenger5
11-04-11, 15:46
Thanks for providing some insight MOGmartin. I'm seeing a drop in vBET translated pages as well, but it is not nearly as dramatic as yours. The difference is around 25% from the previous week. (March 26 - April 1st vs April 2 - April 8)

vBET
12-04-11, 12:54
I will send you detail information via PM in next serveral days...

Please look this article:
technospot.net/blogs/strategy-remove-machine-translated-page-googles-index/

At this moment we do not see your license - please PM license details ASAP. If you are not licensed user, then you do not have paid vBET version. Please send license information.

About the link - it is again recommendation to disable weak automatic translations which can be not understood by human.

We recommend that you do not allow automated translations to get indexed. Automated translations don’t always make sense and they could potentially be viewed as spam. More importantly, the point of making a multilingual website is to reach a larger audience by providing valuable content in several languages. If your users can’t understand an automated translation or if it feels artificial to them, you should ask yourself whether you really want to present this kind of content to them.
Google translations are very good and human readable, so there is no issue with those. So what to make it clear - it recommends to disable ONLY translations which can be not understood by human. We are providing support for people writing in different languages and translations are really good. So this is NOT about Google translations provided by vBET.

I think that in panic and without detailed reading of Google statements we have here lot of misunderstandings. So to make it clear:
1. Translations are OK for Google
2. Google do not like weak automatic translations which are not human readable
3. Having even weak automatic translations is not prohibited, but only not recommended
4. High value automatic translations are OK for Google
5. vBET provides Google translations which have very good quality and are human readable

vBET
12-04-11, 13:15
no probs michal,

its: CONTENT -> Top Landing pages
FILTER: Pages containing this string:

/bg/|/hr/|/cs/|/da/|/nl/|/fi/|/fr/|/de/|/el/|/hi/|/hu/|/it/|/no/|/pl/|/pt/|/ro/|/ru/|/sr/|/sk/|/es/|/sv/|/th/|/tr/|/uk/

...

JUST TO RE STATE AS WELL:
I am NOT SAYING THESE PAGES ARE DE-INDEXED.... They just are not returned in the search results when they used to be.

Thanks I will play with it and check some of our forums.

About re state - as I show you here http://www.vbenterprisetranslator.com/forum/vbet4-general-discussions/1612-auto-translated-pages-vs-google-content-guidelines-3.html#post7908 Google returns your page in search results. I think that you can have less searching entries because of changes in index position, but definitively your pages are still in index and are returned in search results.

We will make some own investigations - also appreciate your future input if you find out something more about situation of your forum in Google index.

AfrikaansAlbanianArabicBelarusianBulgarianCatalanChineseCroatianCzechDanishDutchEnglishEstonianFilipinoFinnishFrenchGalicianGermanGreekHaitian CreoleHebrewHindiHungarianIcelandicIndonesianIrishItalianJapaneseKoreanLatvianLithuanianMacedonianMalayMalteseNorwegianPersianPolishPortugueseRomanianRussianSerbianSlovakSlovenianSpanishSwahiliSwedishTaiwaneseThaiTurkishUkrainianVietnameseWelshYiddish
Languages translations supported by vBET 4.10.1