As some of you know from my previous posts or from Vic’s videos I’ve been working on a “Content Mangler” program. I wanted to write a little bit about what I have in mind with it and how it is designed. I would appreciate some feedback from you on where it’s headed and what you would like to see in such a program.
The objective of the program is to create blog posts that blend keywords, stock phrases, sentences and paragraphs and other base text, such as a PLR article, into a single article. The user will provide up to 10 keywords or phrases. The content and number of stock phrases to insert is determined randomly. The other supplied text can be inserted ‘as is’ or it can be modified by variations of the following ‘manglers’: common misspellings, randomly switching the position of the sentence order in the post or by switch the order of words in a sentence. The user may also choose to bold all keywords or a percentage of keywords. Also, the user may choose to add links for specified words in the text.
The resulting post is not expected to be of high quality but it should have enough quality so that it could easily be confused with poor English writing skills or output from a translator program unless taken to extremes.
The Stock Phrases
The stock phrases used in the program are ones that are commonly found in blog posts and comments and on forum posts. These are mixed and matched in kind of a ‘mad lib’ fashion from between 2 to 10 different parts. So, you might get something like…
“You can expect Losing Weight to provide you with pleasure.”
…the first time and the next you might get…
“It would be a surprise for Diet Food to become a practical alternative.”
The text in blue is from a list of sentence starting stock phrases, the bold is from the keyword list and the green text is from the second list of stock phrases.
One of the challenges in the program is to fill the database with enough phrases so that duplication is virtually a non-issue. There are about 25 phrase lists and most of these lists are over 100 entries that this point and I’m still continuing to add to them. Since, according to what I’ve read, Google’s dupe filtering only comes into play when there are more than 12 to 15 consecutive matching words I’ve tried to keep the phrase pieces to 8 words or less.
I’m still thinking about the level of control to provide in this area so feel free to add your input.
One of the problems with using PLR articles in blogs is that you’ll trigger the Google duplicate content filter. The mangling algorithm is designed to prevent this in two ways.
First, by switching the sentence order in the post as a whole you can reduce the likelihood of being filtered. When combined with injecting enough other sentences into the text this seems to satisfy Google that your content is unique enough although it won’t necessarily pass Copyscape.
Next, by switching the order of words in the sentences the of article it almost assures the that text will not be flagged as duplicate content. Surprisingly, the text may still be marginally readable if it isn’t overdone because the human brain will auto-correct word misplacement. Did you notice the reversed words in the first sentence of this paragraph when you first read it?
You can see output examples in the comments on this article: Niche Blog Content Ideas and Experiences
What Can It Be Used For?
Of course, what is generated is poor quality and isn’t what you would want to have for most blogs. However, it can be useful in other ways.
First, it can be used to build unmonetized, keyword rich, link farm blogs that reinforce your money making blogs. This was something that Vic pointed out.
Secondly, the output can be used as a creative springboard to build your own, passable quality, rambling, unique article. If you wanted to take a 300 word PLR and turn it into a 1000 word, Griz worthy, rambler, the addition of the stock phrases and using a slight amount of scrambling makes it easier to do this.
A third use would be to inject misspellings and/or links into an existing PLR article, with or without the additional stock phrases, in the hope this would avoid dupe filtering.
Of course, with a little modification this engine could become a blackhat web site generation tool rather easily. I don’t want to go that route with it. I think some level of manual intervention with the output will produce the best, long lasting, results.
Do you see any other uses? Any thoughts on must-have or like-to-have features? Think it’s an awful idea as I’ve described it. If so, let me know. Having some ideas on how people would like to use it helps me create a better program.
I’ll be looking for about 5 to 10 beta testers for the program in about 2 weeks. If you’re interested in testing the program, let me know by sending me an email using the Contact page. If you don’t have time to test, feel free to leave me any comments or ideas here.