Saturday, January 21, 2006

Automated Content - The Holy Grail?

We've been experimenting with automated content for some time but have yet to find the perfect solution. Getting to the top of search engines is one thing - but having a site that is interesting and converts is most important when the visitors arrive - and these two goals seems to be frequently at odds with each other.

Recently we've tried articlebot for the task. It's a program that rewrites a page of content using a thesaurus and a common phrases database to rewrite the same content in different language. You can generate hundreds of versions of the same content that is more than 30% unique - so in theory, you can create tons of unique content - and some of it will rank highly.

The results? Well, in some circumstances it works - you never know which pages will rank, but some of them certainly do get good ranks. The problem is the content is generally crap. It reads like it was written by someone that is just learning English and is fumbling for words and uses incorrect phrases and grammar all of the time.

Hardly good content to create trust and keep people reading. We gave up on it after a few attempts.

Automated content has to pass the Touring Test - It has to sound like a real, intelligent human being wrote it.

It seems like we'll stick with caching RSS feeds for now but it is not as good at getting to the top.

Comments are welcome.

No comments: