Monday, March 26, 2007

How does Google Reader know which items I actually liked?

This post by Robert Scoble inspired me to write about an idea that's been on my idea for a while now. He comments that "I read more than 29,000 items in the past 30 days with Google Reader". My count (1,600) isn't nearly as impressive, but according to Reader's stats, I've read 90-100% of them...

Is it just me or are those stats useless? Reader doesn't know how many items I've read - just how many times I've pressed "J" on my keyboard!

The way I use Reader is to press "J" repeatedly, skimming over the headlines, and stopping when something catches my interest.

If Google kept stats on the items I spend the most time looking at (rather than just "how many") it might be more useful. This information could be combined with that collected from other people who have similar interests. Then, when I choose "Sort by auto" it could have a better idea about which items may be more likely to interest me.

Who cares about information overload if I can subscribe to 100+ feeds knowing that the items which interest me will always be at the top? Who cares if I won't read them all? It's not about reading everything, it's about reading something interesting.