Feedster

Profile: Feedster

Next Story

Update: Google Video Cracked

Company: Feedster

Founded: March 2003

Status: Founded by Scott Johnson in March 2003. Merged with RSS-Search founded by François Schiettecatte in June 2003. Announced Series A funding on June 2, 2005 led by Selby Venture Partners. Other investors include Omidyar Network, members of the New York Angels, Kevin Hartz, co-founder of Xoom, Joe Kraus, co-founder of Excite and Jotspot, Josh Kopelman, founder of Half.com, Scott Kurnit, founder of About.com, Mark Pincus, founder of Tribe.net and Support.com, and Narendra Rocherolle, founder of Webshots. Link

Location:

Feedster, Inc.
116 New Montgomery Street
Suite 605
San Francisco, CA 94105
Voice: 415-348-9119
Email: info@feedster.com

What is it?

Feedster is one of the original real-time search engines, and has added interesting new services along the way to further evolve the web 2.0. The services we will profile are search, link search and their new user tagging feature. They also have a nascent RSS reader and other services (like “feedpaper” (which we just can’t figure out), job search, and feed search for sites. Most of these other features are available under advanced search and My Feedster.

In their own words, “Feedster is first and foremost a search engine, however unlike a general web search engine like Google, Feedster includes only a certain type of content, called feeds, which provide many advantages.�?

“Feedster is a rapidly growing Internet search engine and advertising network that provides timely and meaningful information to consumers and large Internet sites in need of targeted media. Feedster provides a fresh index across over 8 million feeds several times per hour, adding millions of new documents daily. Feedster benefits from the ways that blogging is changing the Internet’s basic building blocks – from unstructured web pages to structured documents.�?

A slight tangent starts here….

While attending the vertical leap conference on search yesterday (we will post about this event separately), a lot of questions came up regarding “old search” v. “new search”. Old search (the gold standard is Google) prioritizes results based on “relevance”, which is largely determined based on links into the content. Lots of links = high relevance (this is simplified of course). With real-time search (blogging, news, etc.), link analysis breaks because there is not sufficient time for links to materialize and become indexed.

Real-Time search engines like Feedster and Technorati (Profiles here) generally use “freshness” as the determining factor of relevance. The most recent post including your searched keyword or tag is placed first in the results. For Real-Time search to have its “ah-ha!” moment, these services must figure out a better way of adding relevance to results. This can be as simple as putting new content from highly linked blogs higher in search results (although this may tend to “lock in” older blogs), to creating highly complicated algorithms to determine relevancy of a given publisher to the particular content (so boingboing, the most popular blog, wouldn’t necessarily be given higher relevance if they posted on a new biotech company, something they don’t generally cover).

…tangent ends

Feedster Search:

Feedster search is good but not as good as Technorati is today. The reasons: integration of publisher tags into technorati results, and the fact that Technorati adds delicious, furl, buzznet and delicious tags to their results (see here for instance). Feedster doesn’t do this (yet?) and so the results aren’t as useful (same search at Feedster here).

Feedster search is, however, quite useful and has a very clean interface (something Technorati is criticized for). Feedster also has a toggle to show results by “Date” or “Relevance”, which is great, although relevance seems to be powered mostly by keyword counting at this point.

Screen shots of Feedster Search interface and results:



Feedster Link Search:

Feedster also has an excellent feature that shows link coming in to a particular URI. It’s useful for research and ego-searches, and generally to see who’s linking to what. The interface is very similar to general search.

Screen Shots for Feedster Link Search:

Feedster Tagging:

This is the really interesting new thing happening at Feedster. (note our editorial comments on the benefits and perils of user tagging in a recent profile of Celebrity Flicker – in general, you need to find a really good incentive for users to tag (see Delicious and Furl) or you end up with bad results).

Feedster tagging was announced by Scott Rafer, Feedster’s CEO, in a personal blog post on May 19, 2005 (link). The idea is to add a “Tag this” button at the end of a blog post where users can tag the post as they wish. The code to add this is available here (just do a view source on the page). You must be a feedster registered user and go through a captcha process to insert a tag. Here is the process visually (last one is the code needed to insert it into a blog):

Feedster Tagging Screen Shots:

This is an interesting and ongoing experiment with user tagging. However, currently the tagging results are useful for only one thing: viewing all tags for a post once you’ve added a tag. For now, the results are not integrated into Feedster search or anywhere else, and so there is almost no incentive for a user to tag content in any meaningful way. Thus, tags like “asshat” and “qrp2″ appear within the results. Not good.

I spoke to Scott Rafer about this issue yesterday and he says that they will roll out additional functionality in the future, including support for scuttle, the open source “delicious” bookmarking service. That will certainly help. But for now, the experiment shows that user tagging absulutely requires incentives or else users will not tag, and/or will not tag appropriately.

We’ve added the “tag this” feature to this post as an experiment to see what kind of data is generated. Have at it!

Management:

Scott Rafer – President and CEO
Chris Redlitz – VP Sales and Marketing
J. Scott Johnson – Co-Founder & CTO
Fran̤ois Schiettecatte РCo-Founder & Chief Scientist
Oren Michels – VP Engineering
Link

Links:

Fact Sheet

About
Corporate Blog
News
PR
Review, opinions
Financing
Feed of the day
Developer wiki
Help
Scott Rafer (CEO) blog (read this for his thoughts on the tagging product)
fusion94org on feedster tagging
Brian Del Vecchio on feedster tagging
Maruo Cherubini on feedster tagging
Fredonsomething on feedster tagging
PodTech interview with Feedster Founder

Tags: , , , , , ,

blog comments powered by Disqus