>>I was told to go for 40% + difference which is easy if you are using Markov< <
Sounds good to me.

>>I have WH sites and do not want to risk my adsense account, and have not found anything else that converts.< <
There are all sorts of affiliate programs out there J. Any of them will pay out more than adcents. Take some time to check out Commission Junction and Shareasale. Find a couple of programs that semi match your content and plop up some banners. You can even write an original article to be the intro to your markoved shite.

>>I am currently working on a spinner which takes content from a product database and creates “reviews” of each product. I am mixing up as much as I can, using at least 5 variant of each sentence, changing order of sentences / blocks etc…< <
Sounds perfect for affil marketing. Give it a shot.

>>I plan to gen up 1k + posts on each blog, making 5k posts in total, but I really don’t know if the level of variance I can achieve will be good enough for that number of posts<<
I think it will be enough variance if you make your blog so that it only shows one post per page. That way the viewers won’t get (too) confused.

Nope, haven’t looked at TNX. I will though. Thanks.

And, yes, I understand the synonymizer making perfect output. I like it for sites where there will be real viewers.

Good comments. Keep us posted on how it’s working out.

~dink

Thanks for taking the time to reply.

I was told to go for 40% + difference which is easy if you are using Markov

When using Markov I usually find 10 decent articles to seed from, 300 - 500 words each, so 3-5k words in total. From that I will gen maybe 1000 blog posts.

The problem I have with Markov is monetisation… I have WH sites and do not want to risk my adsense account, and have not found anything else that converts. Also I do not have the infrastructure to do it properly in terms of servers / ip’s / anonymity…

So I have not used Markov for a while and am currently re-writing using synonyms, which I enter manually.

Needless to say, editing synonyms manually is very time consuming so it is only gonna work if I can gen a lot of content from it… hence my worries about the level of difference.

I am currently working on a spinner which takes content from a product database and creates “reviews” of each product. I am mixing up as much as I can, using at least 5 variant of each sentence, changing order of sentences / blocks etc…

The spinner will output RSS which I will scrape back into 5 blogs using WP-Omatic:

http://devthought.com/wp-o-matic-the-wordpress-rss-agreggator/ )

I plan to gen up 1k + posts on each blog, making 5k posts in total, but I really don’t know if the level of variance I can achieve will be good enough for that number of posts….

If they don’t get any traffic I will whore the pages out on TNX.net to buy links back into other sites… have you looked at TNX.net?

My min level of difference varies depending on what I want the content to do. If I’m just posting the content to get spider action, I’ll just put in without much editing at all.

If I’m using the content within a page generator, especially one that uses the markov() function, I’ll add about 20-25% original stuff. Note that I use a whole lot of reference material for the markov to compare with. My minimum no. of words in the reference is 2k. It goes up from there depending on the niche.

If I’m making a post on a blog that I own, then I use at least 50% new content. Mixing sentences and paragraphs so that it is different.

I haven’t used any tools that tell me what percentage of difference there is between two texts. I’m sure they are out there. Maybe copyscape does that. Don’t know. I just fly by the seat of my chair.

BTW, welcome. Feel free to ask questions, make pointed comments, suggestions, or just general hell raising. This blog is as much for you as it is for me.

What is the minimum level of difference you aim for and can you recommend tools to check this (difference)?