Feed Subscriptions
Rolling Links
- Flowing Data Exploring all things data!
- Web Science Research Initiative
- Mashable The Social Networking Blog
- Science Blogs
- Kottke: Fine Hypertext Products
As we wind down our beta period for Recycled Canvas we’re in the process of preparing an internal report of some of the results. We’re planning on getting the Recycled Canvas 1.0 release out later this week. Until then, here are some highlights from the experiment that we’re putting in our report.
There is no denying we got a lot of data.

With some pretty decent download stats.

Highlight: A Single User Path
In this example we can observe a single use page. The user comes in from the upper left of the archive page. They pause on the second article, and proceed to the third. They then quickly return to the second article, which they scan for some time (without conversion) as they exit the page to the top.

Highlight: Reading Patterns
Reading patterns are difficult, but not impossible, to interpret. With careful filtering certain patterns and key sections of an article can be identified. It is also easy to identify basic user behavior. In this case the article demonstrates large degrees of ‘drop-off’ - the rate at which a user abandons an article. Very few users continued reading this article below the fold.

Highlight: Interpreting Results
From left to right, these results suggest things like:
- Sidebar Placement
Sidebar placement on the left, although non traditional in blog design encourages short distance travels between headlines and sidebar navigation. - Resting
Whitespace near the edge of the page often serves as a resting spot for the mouse for some percentage of readers who like to get the mouse ‘out of the way’ to help concentration on a page. - Action Identification
Here we can identify a specific action, in this case feed subscription, along a use path. High percentages of such users (possibly indicating detailed interest in the page) produce this path.

Highlight: Image Scanning
There is little doubt that adding images to the beginning of posts serves as a strong attractor. Here we see a number of users move back and forth between the article text above and below and the image. Hover times for the image are remarkably large, even with the relative simplicity of it.

That’s all for now. More to follow in the next two weeks.









