Radar Begins To Raise Its Head
We had a board meeting yesterday at Radar Networks.
One of the 'truths' of the VC business is the "Oh Shit" board meeting. Generally, this is the board meeting where you find out the underlying basis for your investment was flawed. John Jarve at Menlo Ventures holds the record. As I understand it, he fought resistance from his partners to even get the UUnet investment done early in the Internet boom, AND at the first board meeting after the close the Company was already out of money! (I may have these facts wrong, but it still makes a good set-up.) He had to come back to the well after the Oh Shit Board meeting. UUnet ended up being a monster hit for John and for Menlo.
There is another "Oh Shit" board meeting. This is the one where the idea begins to show real form and you finally see the materialization of your investment thesis. Software is a visual medium. Imagine describing Visicalc to the Apple II user in 1979 while it was in development. "Well you see, it is this non-procedural, general purpose, non-command line thing that solves equations." Doesn't sound like a Killer App.
Yesterday the team at Radar took many of the pieces they have been building and assembled their first real, usable demonstration of the platform they have built. They have spent the better part of a year building a semantic applications platform. Now that the foundation is built, the scaffolding is quickly being raised. Nova's feeling bullish enough about this that he has started to open up a little bit about Radar. At least he is telling the World what it isn't. Dan Farber also picked up on this to give it some perspective.
Actually to all you VCs who have asked me about why I am interested in the Semantic Web as a theme and what I have seen that's interesting, I would suggest you read his post. It does a nice job of laying out a larger framework for semantic applications in the course of beginning to define the sandbox in which Radar will play. He does a nice job of trying to de-hype the Web 2/3/4 sequencing into more tangible and technical distinctions.
I share in your excitement about Radar, in part because their natural language work sets them apart from the 'Get Google' pack. As I'm preaching to any choir that will listen, the Internet as we know it *must* grow smarter in the next few years and begin to learn from our behavior. Radar gets that. Check out the brief profile of the company in the inaugural Guidewire Report - http://www.guidewiregroup.com/site/pdf/tgr/sampleIssue.pdf. And keep us posted!
Posted by: Carla Thompson | February 14, 2007 at 03:07 PM
Awesome post. We've stopped doing the VC pitches until our product is done, it just doesn't make sense. The vision and the plan are there, but without the product they can't fully grasp the power of what we're doing. On-demand software for the public relations industry... "great market, sounds like a great product, but I don't get it fully". Honestly, how could a VC get it, unless they were PR professionals in the past. It's like VisiCalc, you couldn't see it as a killer app until, well, you saw the actual killer app. I firmly believe that unless you have a proven track record or very personal relationship with the VC, don't expect to raise institutional money without the product. It doesn't hurt to start talking and building a relationship. The semantic web makes sense as well. Like you said in an earlier post: solve problems. The semantic web does this by giving the user a more intelligent response. I'll be watching out for Radar. Take care.
-Jason L. Baptiste
Posted by: Jason L. Baptiste | February 14, 2007 at 07:22 PM
Wow, I'm in way over my head with this semantic web stuff but if it is what I think it is (after reading the Radar piece) then I can understand the 'oh shit!' part of it. Meta information on information itself to enable better sorting and interpretation. Am I completely missing it?
By profession I am an online marketing person which means I'm already trying to get a grip on how this affects my world. 2010 ain't so far away. Very interesting to say the least...
Posted by: Martin Edic | February 15, 2007 at 11:57 AM
Actually, Martin, you have the same reaction that 99.999% of the World does. You were polite, but the more direct question is Why Do I Need This Semantic Web Stuff?
No one does. Just like no one *needs* HTTP or Java or RDBMSs. They are all a means to an end. The end (as in the Application) is what matters. There is nothing magic here. It is just a more flexible way to manage information. The *reason* for wanting flexibility is the Application. The Application could be Search or it could be CRM, i.e., it could be anything.
Posted by: Peter Rip | February 15, 2007 at 01:22 PM
Knowledge Management systems have for years tried to encourage people to code meta data. They don't. Time is a disincentive.
Digg and its clones succeed in that tagged content is motivated by social status within the community. Or Calcanis-esqe monetary rewards. Those third parties are supplying meta data to validate the content.
So while I (maybe) understand the goals of a semantic web application, I am curious about incentives to individuals to provide the structured data. Microformats are cool, yet not widely used. Perhaps the average human can't easily type a GMT adjusted ANSI date time embedded in SPAN tags, etc, etc, etc...
I hope Radar Networks works. My gut is very skeptical. Per previous comments nobody "needs" it, yet we would benefit if the incentives kept quality above spam contributions.
Posted by: eschipul | February 18, 2007 at 02:21 PM