Feedscrub is a new, useful little web app I happened to stumble across this week that promises to help those of us who cling to our RSS readers in favor of the river of information or the live web save time.
I’ve done a bit of socializing lately here in Dallas at the area gatherings, and I routinely have introduced myself as semi-well known blogger from my former job, and the most common thing I hear is that regular folks (that is, folks who don’t do the blogging thing for a living) don’t obsessively read every post from high traffic blogs, and with blogs that routinely put out 20-40 posts a day, as many do, they find that they just don’t have the time to bother with it.
FeedScrub is an attempt to alleviate that pressure.
It functions simply enough, and integrates with your existing feed reader, including my favorite, Google Reader. You simply take your high volume feeds (that often contain good information, but also likely a lot of noise as well), enter them into the system, and replace your subscription in your feed reader with the new RSS feed given to you by FeedScrub.
You then need to engage in a few day period of “training the system,” which consists of simply telling them which items in the feed you like and which ones you don’t.
After a while, it builds up a library based on scanning the words in those posts of generally what you like.
Their screencast on integrating with Google explains in a bit more depth.
Who This is For, and Who This Isn’t For
I’d say that this best benefits the average and casual users of RSS readers. Heavy users and power users of Google Reader and other similar services will become annoyed from a couple limitations in the system – some that can be worked around, and some that I’m not so sure can be fixed.
First of all, the act of creation of a new feed will be annoying to those who not only track how blog posts are shared (like RSSmeme and Readburner), but those who do the sharing and the reading of those shares. It’ll make things more difficult for those up and down the chain, unfortunately, particularly for those who read a shared feed, are interested in subscribing to a feed based on a post that’s been shared, and instead will end up with your custom feed.
The other thing that’s annoying about the system – though admittedly this is probably a temporary problem – is the limitation FeedScrub’s made of 3 feeds per account. Casual users of RSS probably won’t see this as a problem, but power users will quickly use that limit up in the first five minutes.
As the service grows, though, I imagine that will be opened up much wider.
Aside from these issues that leap out at me as I’ve played with the system, it’s actually surprisingly accurate on what it is I’m interested in filtering out of a feed. The first one I stuck in there was the PE Hub full news feed. I’m essentially interested in whatever news they have in that feed, if there’s a tech angle to it at all – they do M&A’s and all sorts of news for large business transactions, so it’s not uncommon to see news of a restaurant chain or something pop up in there.
I trained the system just based on the first 20 items, and it seemed to get a pretty firm grip on the type of stories I was interested in, though, which means that the mechanics of this service are pretty well in place.
How Can You Try It Out
FeedScrub is currently in closed beta, but I was able to get my hands on a whopping 500 invites, so grab them while they’re hot.
Simply head on over to the service, and use the invite code “rizzn” to get in. If we run out, leave a comment here, and I’ll see if we can’t scrape up a few more somewhere.