Technorati’s VC Infusion

Áine on August 28th, 2004 filed in Blog On

Interesting entry at Dave Sifry’s blog about the 6.5m venture capital infusion into Technorati. Sifry is the CEO of Technorati. Most of the comments on that entry are typical back-patting… until I scrolled down to this comment:

The VC is the most interesting thing about Technorati. I, for one, would give you my left toe to know what on earth is in that business plan. Is there a business model behind Technorati beyond a lackluster search engine, a badly tuned index, and tracking interlinkages? What’s the master plan?

The problem with the dot com era wasn’t that we all gave too much of our attention to VC, it was that we didn’t give enough. No one questioned the logic behind investments, and for most it seemed the important thing was getting in on the deals.

There are questions I’d like to know the answer to, and I’m sure many others would too. How many queries per second are you handling in your DB that doesn’t even respond half the time and can take up to nine seconds to return responses? Why are you using 60 servers? That really seems off even for incredible amounts of traffic. It seems like all this money breeds a company culture that throws money at design problems instead of innovating and optimizing.

How long is this 6.5M going to last your firm? You’ve got a big data set, a bright set of developers and more money than you could possibly need - what are you doing?

If Technorati has some sort of hidden plan to make it a relevant long-term project in some sort of way let us know. But if all you have in store for us is more of the same then let us know so we can throw up our hands at just another dot com style VC misadventure.
Posted by: John Doe at August 26, 2004 06:35 PM

All good points, I thought. Much of the time, the server speed is lackluster at best. The profile, while it’s a nice idea, doesn’t currently behave the way the settings are supposed to make it behave (my settings wouldn’t stick, anyway). There’s also a small problem of repetitive links, I’ve found, and scrolling through five pages of the same exact links doesn’t give me much useful information, other than that the Technorati database is quickly filling up with garbage. This results in a bloated database that doesn’t reflect a true picture of the blogosphere. A blog may show 15,000+ incoming links, but many of them may be from it’s parent domain or from every single page of a blog (because the link is in a recurring sidebar). Technorati doesn’t seem to do anything about eliminating duplicates. That’s not an accurate link count or an indication of how popular the blog may or may not be. Is ScottWater more linked to than Slashdot? I don’t think so.

I also wonder how they’re handling the influx of porn blogs and linkfarms springing up all over, almost all of which ping weblogs.com. Do they have filters or a mechanism in place to clean their data set of all that, or do they just not care that they are giving these “blogs” (which aren’t real blogs) publicity or pagerank?

Technorati Tags: Blog On

2 Responses to “Technorati’s VC Infusion”

David Sifry Says:
August 28th, 2004 at 3:10 am
Thanks for the comments, and thanks for your feedback on Technorati’s performance. You make good points, and we’re working really hard to address many of them. First off, lthe really important number is not the number of links, but the number of sources that are linking to you. As for links from sidebars or from parent domains, we’re working on filtering those out, both by working with blog publishers, and with a variety of filtering techniques. We’re also dealing with Link spammers and blog spammers (people who create lots of blogs that are automated and have no original content) in a variety of ways, more to come. As for the porn weblogs, there are a number of them out there, and we’re not going to push them down artificially so long as they are blogs. This ends up being something of a judgement call, but we’re working hard to (a) keep the service up and running, (b) make it fast, (c) make sure it is accurate, and (d) pull out unnecessary noise.

Thanks again for your feedback, please do keep giving us feedback, we really care about what you think, and we’re working on keeping the service up to your expectations.

Dave
Frank Kelly Says:
August 28th, 2004 at 2:45 pm
If there’s a filtering solution, great, but it takes human editors to make judgment calls. DMOZ does a good job of blocking links farms and affinity sites, and they rigorously partition “adult content.” Where DMOZ goes haywire is keeping up with content that moved to a new domain or rightly ought to be indexed under several categories.

I’m new to the weblog world, but it seems to me that high quality content is scarce.