Personalization of rating scores

Earlier today we opened up our rating widget to the world. If you have a blog, go ahead and get it here.

One additional important feature we're releasing today is the addition of the personalization algorithm to our rating widget.

From now on, whenever we serve the ratings widget to a reader, we look at his/her rating history, find like-minded people automatically, and adjust the rating scores accordingly. This is what the widget looks like with personalization:

In other words – two people might be looking at the same blog post, but each will see vastly different scores based on each one’s personal rating history. The more each person rates, the better our recommendations will be specifically for him/her.

We think this is really important for two reasons:

1. When consuming content, personal tastes are waaaay more important than averages. Consider this example - The movie Pulp Fiction is probably loved and hated by an equal number of viewers. On average it would be rated say 3.5 stars (of 5). For potential Pulp Fiction lovers, this would be too average to rise above the noise. Many movies would probably fall in the 3ish range and therefore would not be recommended.
   On the other hand, potential Pulp Fiction haters might consider seeing the movie because 3.5 is after all a positive score.
   So with average ratings, no one really gets a meaningful recommendation experience. If your rating widget is limited to averages - it's definitely time to switch to outbrain.
2. The second reason we think personalization is so important is that it finally gives your readers a true personal incentive to rate stuff (and rate things honestly), as they get tangible, long term value from building a rating history.

Again - the outbrain rating widget is available, for free, here. No registration required.