The RSS Blog

Two months ago, I wrote a comprehensive blog entry on the State of Blogosphere Search. Since I wrote that article, I have noticed a considerable decline in the quality of results returned by the blogosphere search engines and wanted to update everybody on what I'm seeing. Let's begin with the infrastructure that is in place for making the World Live Web happen. Two primary issues are negatively affecting the ability of blog search engines to quickly report accurate conversations in the Blogosphere; the broken ping infrastructure and splogs.

Blogosphere Ping

In my last report, I noted that many of the blogosphere search engines were failing to pick up new posts via the pinging infrastructure. With a little help from FeedBurner, I was able to better Technorati's response time in indexing my posts. The problem related to Ping-o-matic. When PingShot was configured to ping Ping-o-matic, it would withhold its ping of Technorati, assuming that Ping-o-matic was pinging Technorati on its behalf. Unfortunately, Ping-o-matic does not appear to work. By removing Ping-o-matic from the list of ping services that PingShot would hit, FeedBurner began pinging Technorati and my posts are now picked up in reasonable time. In other words, Technorati pings work for me, because I was able to figure out what I was doing wrong with the help of some friends. Unfortunately, most bloggers wouldn't know what's broke and what to fix.

I still have little luck with PubSub, which seems to miss the vast majority of my posts, even though PingShot pings PubSub with every post. Google blog search is now picking up most all, if not all of my blog posts in reasonable time. Blogdigger is falling behind one or more days at a time. BlogPulse is usually behind, but not more than 24 hours. Most of Feedster's functionality was recently broken, then temporarily dropped, so I don't know how their index is doing anymore.

Overall, the blogosphere ping has improved and the improvements lie with FeedBurner and Google blog search. Everybody else has stagnated. On the downside, 75% of all blogosphere pings are now spings (a.k.a. pings from splogs). Which brings us to our second topic; splogs.

Splogs

Searching almost any blogosphere search engine returns splogs. Although a lot of progress has been made, the increasing number of splogs is making this difficult. Blogspot has been able to reduce the amount of splogs, but the sploggers have moved to other platforms, including self hosted, which allow them to stay in business, even when caught. Fortunately, when caught, they're quickly removed from the blogosphere search indexes.

I've noticed a considerable improvement at PubSub and Technorati which were splog infested two months ago. Now the results compare with the rest of the industry.  Google blog search, IceRocket and BlogPulse, which were already good at handling splogs, are now even better. Blogdigger also seems to return few splogs.

Now, let me finish with a summary of who's best at blogosphere search.

IceRocket

IceRocket remains the #1 blogosphere search engine. The response time is sub-second, it always works, good splog filtering, index is timely and complete, has tag and link search and has ah-hoc RSS results. Missing is blog profiles. Grade: A.

Google Blog Search

Google blog search compares to IceRocket in every category, but does not provide a tag search. Grade: A-.

Technorati

Technorati has the most complete feature set, but their are too many problems. Response time is mostly sub-second, but not always, getting indexed requires too much effort and the results are often untimely. Grade: B.

BlogPulse

The BlogPulse profiles are their one plus. Other than that, the response is slowing, it often doesn't work, the index is not current, nor is it complete. They don't provide tag search. Grade: C.

Blogdigger

Blogdigger is better than I reported 2 months ago, but it still has a lot of failings. The first of which is that it lacks in functionality and some of the functionality lacks public documentation. Grade D+.

PubSub

PubSub's index is quite incomplete and it remains very difficult (geeky) to use. Grade: D.

Feedster

Feedster has been broken ever since former CTO Scott Johnson left the company. Feedster has promised to bring us something new this month. Grade: F.

Bloglines

Bloglines citations has been broken for quite awhile now and the keyword search is unusable. Grade: F-.

Yahoo! Blog Search

Yahoo!'s blog search remains in Beta and still doesn't have a clean entry point. I'll refrain from grading it until they provide a clean entry point. Without a clean entry point, it's pretty much useless to the average blogger.

Summary

Note, these ratings are my opinion and thus may be incorrect because of my lack of omnipotence. If you have an opinion, then feel free to express it in my comments or link to this blog entry and I'll find you.