Autoblog best practice: whether to fetch partial or full content from feed?

I agree with you,, but do you have a solution in autoblogging way?

Very easy. A properly-made autoblog aggregates scattered content from various sources and makes it available as a solid, cozy, well-structured and constantly updated website (find most popular and toprated things; sort them by tags and categories; use search function etc).

For example, do you use Google image or video search services? Why would you do so if Google doesn't generate any unique content but just aggregates it from various sources and displays it as a picture gallery of a set of video previews? So the good autoblog does almost the same, and usually it does it even better because its owner sorts out the content sources manually and don't relay on keywords like search engines do.

Doesn't work after Panda. You don't need to "make Google happy". Now you have to "make real surfers happy", because the surfers' behavioral is the most important thing for Google nowadays.

Literally a year ago it was very easy to make Google happy paying a few coins to 3rd-World copywriters to fill out your blogs with "unique keyword reach" content. Panda has ruined such an approach on April 24th.

Be honest with yourself, 99% of paid articles are garbage, because they are not intended for the real people. Course if your copywriter's name is not Stephen King ; ) An ability to write really good articles that will keep the surfers glued to your site is a very rare gift. For example I can't write anything that will touch the people and make them read my texts for hours (even when I write them in my mother language). If you can do that, I sincerely envy to your skills!

The smart people do not use autoblogging nowadays. They use various content automation methods. As I already mentioned somewhere above, Google itself is a good example of a self-populating site which does not produce its own content. Google just automatically aggregates a 3rd-party content from different sources and shows it to you as a well-structured collection of links, images, videos and other sorts of content you are interested in.

Those who got this simple concept make serious money on self-populated sites. Others continue to buy boring and useless "unique" articles which can't make a surfer to stay at their sites even for a minute. They (the sites) get an outstanding bounce rate and end up in the "Mariana Trench" of the Google SERP's.

It was all about switching Google to the so-called behavioral model. Personally I think it was a good and wise move.

Since the AI is not invented yet, the search engines can't estimate the quality of your content. In other words, Google's algo is still unable to read every article on the Internet and conclude something like: "Yeah, that was a real good sh*t, but I would add a bit more drama to the second act where Senior Pedro finds out he is the lost son of Donna Carmelita!"

So the only proper way estimate the quality of a Web resource is to let the surfer to decide which site is better. Fortunately now its very easy to watch for the surfers using such a big arsenal of spy tools like Google toolbar, Chrome browser and that countless number of Google Analytics scripts on almost every site.

It is exactly what Panda was about. And yes, I really love it!

P.S. The only new update what did hit my networks was Penguin, which intended to fight so-called "link spam" when all those old good link exchange SEO techniques suddenly become useless and even harmful. However that's not fatal and there are proven ways to recover the sites penalized for "unnatural linking"