Content generation using Google translate - does it work?

madoctopus

Supreme Member
Joined
Apr 4, 2010
Messages
1,387
Reaction score
3,707
You know that technique where you steal somebody's content and translate it multiple times (e.g. en>de>fr>ch>dk>it>de>en) and you get partially different (and pretty unreadable) content? Does that work? Does google figure it out or it just thinks it is unique content?

I know it worked some while ago but haven't heard about the technique lately. I was wondering if it still works.
 
Do you mean to pass copyscape? If you do I haven't had any success with it.
 
nah its not ideal. You have to be real careful about copying brands and names into the text. I gave it a go a while back and didn't have any real luck. The text turns out a bit like it has been written by someone who doesn't have a full grip of English. Content is still king and translating isn't going to produce good content in my opinion.
 
nah its not ideal. You have to be real careful about copying brands and names into the text. I gave it a go a while back and didn't have any real luck. The text turns out a bit like it has been written by someone who doesn't have a full grip of English. Content is still king and translating isn't going to produce good content in my opinion.

True that.. True that
 
I don't need it to pass copyscape as I will use it for support blogs / web 2.0 blog sites to build backlinks. I just need it to get indexed. No chances of brands and names being copied.
 
I've been playing with this technique a bit lately myself. Preliminary, very preliminary, indications are that it works well for low-level sites that will be used only to send link juice up the chain.

Cool5now wrote "The text turns out a bit like it has been written by someone who doesn't have a full grip of English."

That is an accurate statement. But there are, I think, ways around it.

What I've tried is to introduce the site/blog/etc. with a statement by the "owner" admitting that he "doesn't have a gull grip of English" but that he thinks the material is important and perhaps somebody who agrees could help him translate it better.
 
I would strongly advise against it, at least if you want to use the articles to get readers to click through to your money/landing pages . With each translation it becomes more and more garbage.
 
That all depends on what you're gonna do with it. ;) I'd say use it if you're trying to index a site. Why not? If you're trying to do article marketing and trying to pursuade people to buy stuff though, like Peter2002 said,it's prolly not a good idea. I've used this technique for an adsense site, and have had much success with it. Of course there isn't as much traffic in the other language as there is with english...but that's okay. Think about it...a lot of WP sites or just sites in general have a translator plugin (that uses Go0gle or some other translator) to do exactly what you're doing. Visitors still read it. You can read broken english can't you? You understand and can fill in where the translator fails, right? It's natural for the human brain to fill in the gaps... Oh yes, and here's a little secret that not a lot of IMer's take advantage of when it comes to making money online:

Adsense comes in many languages!
 
i have an autoblog on that concept
its getting 200 - 300 visitors from google daily making me ~ $2 daily.

although content steadily grows, usernumber from google stays the same.
 
It won't work, since there will be translation mistakes etc
 
True that.. True that


It fails on both counts. It rarely passes copyscpae (so no posting on high PR sites for PR improvements) and it's almost always gibberish, so you don't get a lot in the way of traffic from it either.

Have a look at this shite from "mage" site I made a while bacl (before I packed it in - biggest waste of a thousand bucks I ever bought)

Myself, I played ordinarily for threesome years, then as Blizzard prefabricated the mettlesome simpler and simpler I necessary the contestant contest of, prototypal digit enclosing then multi enclosing to ready me filled in the game. It was effort to the initiate with digit case that I could endeavor internet poker, center to penalization and be on the ambulatory to my wife, whilst closing a heroic. One case at a instance was no where nearby substantial enough. If Blizzard illegal multiboxing (and they won't) I would yield Warcraft for more substance

It was supposed to say this.

For me, as the game has become simpler over the years, I have looked to increase my challenge. Multi boxing keeps me occupied in the game in a way that controlling one character simply cannot do any more. It's all too easy. It's fair to say that before multi boxing, I could play internet poker, ‘phone my wife and listen to music and still finish my heroic. One character at a time does not occupy my attention. If multiboxing was was banned (and it won't be) that I would lose interest in Warcraft.

It had had only ONE translation on it (English to Spanish - then back to English)

I tried submitting other crap like this to decent article sites. "Go" accepted at first, but deleted it a few weeks later. Ezine and others were far less toleraant (Ezine suspended the account for 24 hours)

IMO - this technique is crap. Googles bots use a syntax and logic checker like Micoshits "Word". You know all the green and red underlines that tell you a sentence doesn't have a verb, or it's a fragment.

It's a mistake to think that a bot crawls for unique words only - regardless of whether it actually makes any sense or is readable or not.

That might have been the case 10 years ago. They are far more sophisticated now, and this sort of "dogs steam" gets chucked straight into the waste disposal.

Scritty
 
Last edited:
1) Open Google advanced search

2) Enter keyword

3) Select foreign language

4) Select filetype (doc is normally good)

5) Search, and find foreign documents that Google has not indexed in English.

6) Translate them into English, clean up the grammar if necessary, and there you have your content.
 
I belive it actually might be pretty good. I translated a text to Norwegian, then German then English again. According to dupecop I had a 35% unique article and I could still read the entire text without any huge problems.
 
What I want is pages that get indexed and stay indexed not pages that you can read easy. I want this only for link building. Those sites would probably get 2-3 visitors/day. Maybe not even that.
 
What I want is pages that get indexed and stay indexed not pages that you can read easy. I want this only for link building. Those sites would probably get 2-3 visitors/day. Maybe not even that.

I wouldn't go the 'multiple translation' route... even not for the purpose you are describing. It's better to use the method "Micallef" mentioned in his post. You'll get unique content and the translation makes way more sense because you translate the content only once.
 
You know that technique where you steal somebody's content and translate it multiple times (e.g. en>de>fr>ch>dk>it>de>en) and you get partially different (and pretty unreadable) content? Does that work? Does google figure it out or it just thinks it is unique content?

I know it worked some while ago but haven't heard about the technique lately. I was wondering if it still works.

Actually, it depends on what language you choose. Pick one from the same family of languages. E.g. if you want an English output you would go for some of the mid/nothern european languages.
 
French to English works pretty good with this method. Still need to do a bit of editing (If you want it to pass CopyScape), but overall a quick and easy way to save some $ on content outsourcing for those starting out.
 
French to English or English to Swedish and Swedish to English work very nice :) Im using on my autoblog :)
 
I've now developed a tool with a built in scraper which grabs content in another language, extracts all the good quality sentences, filters out all the rubbish and auto-translates the content into English.
 
Back
Top