Good ol' Rand has a bit to say about unique content. Ironically, I'm going to share some of what he has to say (but I'll link/attribute him, and thus lift him up in the eyes of the internet!).
If
For SEO purposes, search engines like to filter out what they view as duplicative content, things that are exactly the same. They never want to show you a set of results where result two, three, four, and five are all exactly the same article or are essentially the same three paragraphs repeated with the same photos embedded in them. ...
He then breaks down 4 big "must get" points to help ensure your content is unique and relevant and "get-found-worthy." First, the unique content refers to the actual content on the page, not the "debris" on the side, in the form of ads or footers, etc.
The first thing that I always like to talk about when we get into a discussion of content uniqueness is that content, when we talk about the content that the engines are considering for this, we're referring only to the unique material on a page. That excludes navigation, ads, footers, sidebars, etc. ...
Second, you can actually kill your own page by citing yourself too much!
Number two, uniqueness applies to both internal and external sources. Copying either one can be trouble. It could be that these are other pages on your site and these are other pages somewhere else on the web where this content exists, and you're taking from those and putting those pieces on your site. That can be a problem in either of those cases.
There are some ways to get around this, including using the "canonical" tag in your html scripting when you do cite often from your own website.
You would be wise in these situations to use something like a rel=canonical. Or if you're consolidating pages after a big site move or a re-architecturing, something like that, a 301 is proper. But you should also be aware that this can happen from external stuff. ...
Third: Uniqueness is not so unique as VALUE is.
Number three, uniqueness alone, some people get lost in the minutiae of the rules around SEO, the rules around search engines and they think, "Well, this content exists nowhere else on the web. So I just took someone else's and I changed all the words." You have technically provided unique content, but you have not provided unique value. Unique value is a very different thing. What I mean when I say "unique value" and what the search engines would like you to do and are building algorithms around is providing value that no other sources, no other sites on the web are specifically providing. That could mean that you take a look at the visitor's intent, the searcher's intent or your customer's intent and you say, "Hey, I'm going to answer each of these things that this person is trying to achieve." ...
Finally, there is no magical "uniqueness percentage."
Four, there's this imagination that exists in the minds of folks in the SEO field, and has for a long time, that there must be some mythical percentage. If over here, "Oh, this is 100% duplicate and this is 0% duplicate, 100% unique and this is the 50/50 mark, there must be some imaginary, magical, if I just get to like right here at 41%, that's the number. Therefore I'm going to create a huge website and all my pages just have to hit that 47% mark." That is dead wrong. Just totally wrong. There's nothing like this.
The algorithms that you might imagine are so much more sophisticated than an exact percentile of what is and isn't duplicate, even when it comes to just studying the content in here. That specific percentage doesn't exist. They use such a vast array of inputs. (
All in all, great stuff from the SEOMoz folks, and worth your time!