This is indeed interesting topic, as today with all different ways of gathering info it really is hard to produce 100% original content. I mean especially when collecting from RSS feeds.

That is what I do, I display headline, link source and small 1 or 2 sentence intro for a news story. I collect from like 20 sites.
I do have a lot of *other* content on my web which is 100% original, but this news inclusion will hopefully give me traffic as people can read daily news (its parts to be exact) from single source.
This is outside US country.

Ways I *try* to differentiate my content is to have dynamic title, description and keywords made upon that short info. Otherwise it's totally duplicated, but I hope/think Google won't penalize me for this as its only a very short bit of info.
So according to a duplicate algorithm my specific news story page is around 15% the same as the source.
Source news article is probabbly 10 paragraphs long and I only use one.

My site is still in development phase and not yet launched, so can't really guarantee for results. I just presented my subjective thinking : )

which is...?

I mean, your example is ok, but how does that answer my question?

Sorry Emoweb, I know I didn't provide any answers, just the similarity of problems.
There is no definitive answer to this one, we'll just have to see what Google decides :(

But I'd suggest this:
- differentiate the job page as much as possible from original, include different meta tags for example
- add some extra unique content to the page, like a brief description of other jobs. Something as "maybe you would also be interested in.." and then add couple of headlines.
I believe this will benefit in the overall "uniqueness"
- there are great online tools that will calculate the similarity between web pages and potential dupe content, try them out and keep changing the page until you get a satisfactory (small) percentage of potential duplicate content
- change certain keywords, for example use "loidging" instead of others, if the original sources don't use them
- if you include pics, don't hotlink them. Have them dynamically saved to your server and change their names.
- perhaps adding a direct user input to the page? Like comments or something? I know it might not be applicable as you are dealing with job offers. But it can create authentic content
- as for bizrate, I noticed they include rel="nofollow" when they link to original content. I bet this is used because of potential duplicates. Not sure how it helps though, but maybe they want to prevent Google seeing directly from where they link :D Makes sense right?
- final advice: create original content that will rank high, and the linked stuff will naturally follow as a direct result of quality. As you mentioned, many sites do it and rank great ! This is not copy-pasting, this is providing user with good sources and easy navigation, it should not be penalized under any circumstances. But let's not forget that Googlebot is a robot : D

Hi emoweb, I am going to do the same thing but with data from 20 sites. The site is supposed to be launched pretty soon.
I have been looking at sites displaying product info and most of them have the same description listed. The pages are indexed and the sites rank well.
Duplicate content is more page related, not site related I assume. Thus, it is important that you make your category pages unique, i.e. use many sources on one page so whole page is unique and use product page as link whores for the category pages.
And I believe it is most important to gain good links as quickly as possible to prove the site is valid and adds value.

Thank you both for your answers.
I also think that bizrate and other sites donĀ“t get penalized because they provide some functionality to what they do. I mean, people can compare prices from different stores and websites in one site. ThatĀ“s usefull, and following my golden rule "if it ads value to the user, then itĀ“s a good idea to do it". But itĀ“s also dupe content. So thatĀ“s why im asking.

I think the best thing is to think what can i do with the content i have from other sites, like i said before, for example create average prices, and other statistic information useful for the users that is not available in the websites where the content comes from.

Does anyone know of other features i could add?

I also think that bizrate and other sites donĀ“t get penalized because they provide some functionality to what they do.

A side note:

Bizrate has been around for quite a few years. They have develped a trust rank with Google, which can help overcome some of the duplicate content issues that a new site may face.

The Shopzilla sites, which includes Bizrate, were rank in the top 50 US web properties for April 2009.

[edited by: MadeWillis at 1:59 pm (utc) on May 22, 2009]

Emoweb,
You post a very relevant and important question. Keep in mind that most search engines 'assume' you have 'white hat' content until you prove otherwise. As more and more content becomes available online, it is my observation that SE duplicate content filters are becoming more forgiving.

Having said that, no one (other than the SE's themselves) knows what mix and critical mass of duplicate content that will trip the algorithms duplicate content filter.

Having built several successful Internet companies, and having done battle with the SE algorithms since their inception, I'm of the mind that harvesting content from other sites for use on your own site only presents a duplicate content problem when you do not take the time to 'artfully' modify the content so that it is truly unique.

Bottom line, when using content from other sites you have to put in the 'work' and develop your own unique content. Given the vast amount of subject matter available on the Internet today, it is not unrealistic/unethical to use content from multiple sites to form your own unique content.